#53

As you like association studies @约瑟夫_拉维尔: Novel Classification of Cardiovascular Disease Subtypes Reveals Associations Between Mortality and Polyunsaturated Fatty Acids: Insights from the United Kingdom Biobank Study 2024

They clustered people into 3 CVD subtypes based on some biomarkers:

Principal component analysis and k-means clustering were used to determine the CVD subtype. Variables included age, body mass index, waist–hip ratio, diastolic blood pressure, systolic blood pressure, total cholesterol, total triglycerides, high-density lipoprotein-cholesterol, apolipoprotein B:apolipoprotein A1, glycated hemoglobin, creatinine, albumin, C-reactive protein, white blood cell count, platelet count, and hemoglobin concentration.
Three distinct CVD subtypes were identified, with cluster 3 characterized by older age, male gender, and low high-density lipoprotein-cholesterol, having the highest risk of mortality. Clusters 2 and 3 had the highest DHA and ω-6/ω-3 ratios, respectively, compared with Cluster 1.
Cluster 1 consisted of the youngest individuals with elevated levels of ApoB:ApoA1, TC, TG, SBP, and DBP. Cluster 2 predominantly comprised females (77.9%) with the lowest BMI, WHR, TG, creatinine, C-reactive protein, WBC, and hemoglobin concentration, while having the highest levels of HDL-cholesterol. Cluster 3 primarily consisted of males (72.7%) who were older, with a higher BMI, WHR, and HbA1c, and a lower TC, HDL-cholesterol, and PLT.

They adjusted for:

  1. sociodemographic factors, including gender and household income (less than £18,000, £18,000 to £30,999, £31,000 to £51,999, £52,000 to £100,000 and greater than £100,000);
  2. socioeconomic status, including Townsend deprivation index;
  3. lifestyle habits, including smoking status (never, former, and current), alcohol status (never, former, and current), and physical activity (low, moderate, and high);
  4. comorbidities, including hypertension and diabetes; 5) drug use, including cholesterol-lowering medication use, antihypertensive drugs use, insulin treatment, and aspirin use.

Results:

image
image1920×931 79.6 KB

In multivariate models, we found significant nonlinear relationships between total PUFAs, ω-3, and DHA and risk of all-cause, CVD, and IHD mortality (all P for nonlinearity < 0.05). There was an inverse L-shaped exposure–response relationship between these PUFAs and risk of all-cause mortality, whereas there was a U-shaped exposure–response relationship for risk of CVD and IHD mortality. In the absence of these PUFAs, risk of the outcome increases. Within a specific range, higher ω-3 and DHA concentrations are associated with a reduced risk of outcomes [hazard ratio (HR) < 1.0]. However, once PUFAs reach a specific threshold level, risk stabilizes and no longer continues to decrease or even increase. Cluster 1 displayed the lowest HR under similar exposure levels in all-cause mortality (Figure 10).
We found significant linear relationships between ω-6 and LA and risk of the studied outcomes (all P for linearity < 0.05). High level of these PUFAs were associated with low risk of all-cause and CVD mortality but were associated with an increased risk of IHD mortality.
Our findings suggest that the mortality rates of ω-3 and DHA with CVD and IHD begin to trend upward at high intakes (fourth quartile array). The possible reason for this is that excess ω-3 and DHA may be predisposing factors for atrial fibrillation

So, even this association study confirms that higher might not be better, especially for DHA. The potential inconsistency with previous studies is, I guess, that this one adjusted for many factors, especially income and education. Rich people eat more fish, and if you don’t account for that, the omega-3 index is just a wealth index. Physical exercise might also help to absorb omega 3, so it’s good that this study adjusted for “physical activity (low, moderate, and high)”.

You also asked “Do these results apply to healthy people?”: here they “constructed a subtype model of cardiovascular patients” “according to patient baseline characteristics”. They identified 3 clusters that I would call “young unhealthy” (1), “healthy” (2, mostly females), and “old unhealthy” (3). The results were similar across these clusters.

2 Likes
2g/day of DHA for 2 years has no impact on cognition or hippocampal volume
#54

As a rule of thumb, Dr. Stanfield has been recommending EPA to DHA in a 2:1 ratio. After looking at the results of these various studies, I may even want to try for a higher ratio of EPA to DHA. It does appear that DHA is the red-headed stepchild of the Omega 3 family. A bit of DHA is probably OK, but you want to limit supplementation to mostly EPA if possible. Eating fish is probably the best way to supplement Omega 3s.

I’m glad that @adssx has brought these facts up about DHA, but it seems like it’s become a crusade now. :wink:

4 Likes
#55

I suspect that it’s partially designed to stimulate opposition. I do that myself sometimes - I will crusade on some issue, because I want people to challenge my idea/findings. Sometimes, if you just casually present something without really pushing it, it just gets ignored. And some ideas/findings are too important to be ignored, like EPA and DHA because it’s pushed so much by influencers and so many people supplement. It’s important to thoroughly discuss it, and hopefully reach some more justified practices with better evidence. For that it is useful to challenge people to present their best opposing arguments and findings. At least that’s how I take it :grin:.

4 Likes
#56

That’s exactly the reason. I’m pushing for people to provide alternative explanations and challenge what started as a question ( Omega 3 makes me depressed: why? ) but what is now, I think, a fact: supplemental DHA is detrimental to your health. So far, no one has provided any valid objections. I hope one of the “influencers” I contacted will get back to me. I have low expectations.

5 Likes
#57

I feel bad for @adssx for pulling together so much information and spending so much time on this, just for people to not read his posts thoroughly and instead just quickly criticize and keep to their pre-conceived notions of the what is right. I hope we can do better in the future.

3 Likes
#58

I think many , like myself, find the information very helpful and will help guide my supplementation strategy.

6 Likes
#59

A few people messaged me directly and told me switched to EPA-only or were considering doing so. That being said: I might be wrong. Mitchell Lee from Ora Biomedical made an intro to Matt Kaeberlein (his cofounder) to discuss EPA vs DHA so I hope Matt will get back to me. It would be good to know his thoughts.

4 Likes
#60

I have not tried to analyse the information in any detail, but I have shifted to taking only EPA. As far as I can tell on a superficial scanning of the information there is clearly an argument against DHA supplementation.

My own personal results with fish oil is that I noticed a benefit from supplementation with a mixture a few years ago. Hence I have now stuck to an EPA only version possibly for a few months now. I have not checked exactly when I changed to EPA.

3 Likes
#61

Chinese paper, just published, not sure if it has any value though: EPA but not DHA improve systemic IR through activating muscle IL-6/AMPK pathway in high-fat diet-fed mice 2025

3 Likes
#62

The last remaining argument in favor of DHA is some association studies such as this one from 2024: Circulating Docosahexaenoic Acid and Risk of All-Cause and Cause-Specific Mortality 2024

Massive association study:

We analyzed data from UK Biobank, which included 117,702 subjects with baseline plasma DHA levels and 12.7 years of follow-up between April 2007 and December 2021. Associations with risk for mortality endpoints were analyzed categorically by quintile of DHA plasma levels.

The results look great, except for CVD mortality (more DHA is not better). The trend is also good, but Q5 is not statistically significantly better than Q3. So it looks like as long as you’re not super low (Q1), you’re good:

image
image2915×1707 302 KB

Digging more into the paper, they adjusted for:

Age (years), biological sex (male, female), race (white, mixed, Asian, black), marital status (living with a partner, not living with a partner, other/unknown), employment status (not employed, low activity work, high activity work, strenuous work), education (college or higher, post high school, high school, less than high school, unknown), Townsend Deprivation Index (continuous measure of relative deprivation), physical activity (very low, low, medium, high, very high), smoking status (never, former, current, unknown), alcohol intake (daily, 3-4x/week, 1-2x/week, 1-3x/month, special occasions, never, unknown), Body Mass Index ([BMI] continuous, kg/m2), prevalent dyslipidemia (yes/no), prevalent hypertension (yes/no), prevalent diabetes (yes/no), physical activity (very low, low, medium, high, unknown), self-rated health (excellent, good, fair, poor, unknown), and total circulating omega-6 polyunsaturated fatty acid (PUFA) levels (continuous, % of total fatty acids).

So, they did not adjust for income. We know that income and omega-3 intake are associated: Omega-3 Long-Chain Polyunsaturated Fatty Acids Intake by Ethnicity, Income, and Education Level in the United States: NHANES 2003–2014 2020. It is a first confounding factor. However, they adjusted for education attainment, which is correlated to income.

Other recent association studies that adjusted for income did not find benefits (or even found detrimental effects) for high-dose DHA:

They also show that serum DHA levels are highly correlated with fish oil supplementation:

image
image1426×568 57.6 KB

There’s a massive healthy user bias here if you don’t adjust for fish oil supplementation: people who take fish oil are likely to be more health conscious. They probably also have a better diet, they might take other supplements, they might go to the doctor more often, they might be wealthier, etc. How come they didn’t adjust for that? Or just show the results stratified by self-reported fish oil use?

I also find the ~18.3% fish oil use in Q1 surprising. It’s self-reported, so maybe these people use a low dose infrequently, but even with low-dose intermittent fish oil supplementation, I would expect most people to be at least in Q2 (omega 3 index around 5%). So maybe other factors impact DHA absorption that are not accounted for? DHA levels might therefore represent something else?

Then, they looked at DHA only. Indeed: “Blood levels of DHA but not EPA were available in ∼25% of the individuals enrolled in the UKBB.” They note:

We focused on DHA in this meta-analysis as it was the only specific omega-3 fatty acid level available in the UKBB because nuclear magnetic resonance technology was not able to reliably measure plasma EPA in this population. Blood levels of DHA but not EPA show strong statistically significant inverse associations with risk of Alzheimer disease. On the other hand, EPA monotherapy has been shown to be effective in reducing risk for major adverse CV events. No similar trials of DHA monotherapy have been undertaken. Levels of EPA+DHA have been shown to be inversely associated with mortality; however, whether EPA or DHA is more strongly associated with improved life expectancy remains uncertain.

So, even them don’t know what’s best between EPA and DHA. I guess that there’s a strong correlation between EPA and DHA levels. What if the best is actually high EPA and average DHA? We just don’t have the data here, and we cannot conclude.

To finish, there’s a risk of bias of the authors. All these association studies are always published by O’Keefe and Harris. O’Keefe is the Chief Medical Officer of Cardiotabs, a nutraceutical company selling omega 3 supplements. Harris works for OmegaQuant, which sells the Omega-3 Index test… What a surprise!

That paper (and other similar ones) is the only argument left in favor of DHA supplementation. And I think for the reasons explained above, it’s not a strong one. Especially when considering all the arguments against.

2 Likes
#63

I’m one of the people who has switched to EPA as a result of his hard work!!! *raises hand enthusiastically

Until I hear a better idea, for now I’m going to supplement with DHA one day per week because I don’t eat fish. The thought is most of you are consuming some dha in your food, and this would help prevent me from having a deficiency.

5 Likes
#64

Yes. I think getting rid of all DHA is a mistake. It seems to be a U-shaped curve for this one. However, EPA seems the more the better. So I’ll be increasing my EPA intake and decreasing but not eliminating DHA.

2 Likes
#65

Just pointing out that they jury may still be out re above

There could be overadjustment bias in those last studies

And given that the current study not only adjusted for education, but ALSO for 4-5 or so other variables that are correlated to income adding income to this study could lead to overadjustment bias here.

1 Like
#67

100 mg/day of DHA is probably enough. But you get more than that with just one portion of salmon per week. So no need to supplement.

4 Likes
#68

Which ones?

I don’t know about overadjustment. In any case the lack of adjustment for income isn’t the worst to me. The worst is the healthy user bias and the lack of presentation of results stratified by fish oil use. What if you don’t have any significant trends anymore when you look at fish oil users only or non fish oil users only? How come they didn’t present that data?

#69

Matt Kaerberlein’s answer:

I don’t really disagree with your position, although I don’t find the evidence against DHA to be all that convincing. Having said that, I haven’t closely read all the latest literature in this space, mostly because of time constraints and also because I get frustrated at how poorly many of these studies are designed and/or analyzed.
One of the biggest limitations to most studies in this area that I’ve seen is that they don’t actually measure omega-3’s either before or after supplementation. It’s tough to draw any conclusions - other than that population level supplementation probably doesn’t have big effects either way - in the absence of actually measuring biomarkers to see where folks started and where they ended up. If you know of good studies that did this, let me know.
Obviously, the relationship between omega-index or specific omega-3’s and health outcomes is extremely complicated and impacted by genetics and other environmental (diet, exercise, etc.) factors, likely as well as whether the source comes from diet versus supplements. Even brand of supplement may matter…
I do agree the evidence for benefit from supplementation tilts toward EPA for many outcomes, at least for now. The one I take is 3:1 EPA:DHA.
Honestly, like most supplements, I think the real answer is that supplementation is unlikely to move the needle much for most people. Especially if you are eating a crappy diet and don’t get other lifestyle factors dialed in. How much does it help if you do practice an otherwise healthy lifestyle? Hard to say, but for now it makes sense to me to get your omega-index up near the range is that most correlated with better health outcomes across populations.

I feel like he didn’t really read my email :sweat_smile: (at least he’s honest: “Having said that, I haven’t closely read all the latest literature in this space”), but he seems to directionally agree (he takes 75% EPA).

7 Likes
Predicting Alzheimers & Dementia (and minimizing risk)
#70

Actually, checking this, many studies looked at the baseline levels and it seems that, unfortunately, supplementation is worse in people with low baseline levels!

VITAL-DEP: Effect of Long-term Supplementation With Marine Omega-3 Fatty Acids vs Placebo on Risk of Depression or Clinically Relevant Depressive Symptoms and on Change in Mood Scores

image
image1920×1723 217 KB

People with low omega 3 levels were worse off with supplementation!

This was also noted by @Neo regarding another study: “The best responders were those with higher (aggregate) initial levels” Omega 3 makes me depressed: why? - #80 by Neo

So it’s even worse than I thought…

2 Likes
#71

It looks like there’s a big sex based difference? There seems little effect on men? Confidence intervals for women is completely within the favoring placebo, strong effect; for men, of course straddling neutral. The thing that’s interesting here is that if you eat fish, skip fish oil supplements.

However, it is also important to note: this is in regards to depression, depressive symptoms, mood disturbances. OK, but not everyone experiences such effects from EPA/DHA. On a population level, prevalence is relevant. And what does that mean? Does it affect brain health only in those who experience such symptoms from supplementing and those who don’t experience symptoms are unaffected?

I don’t supplement with DHA, but do modestly EPA (3x week, 500mg EPA). So I don’t know if I would get depressed if I took DHA. But if EPA supplementation also causes this effect, at least in those with low baseline levels, then it becomes a question of dosage - at what dose is this noticeable. FWIW, I don’t have any depressive symptoms or mood disturbances, and never had them. My modest EPA supplementation hasn’t changed that over the span of 2 years, which is how long I’ve supplemented. But I’m also male, so…

2 Likes
#72

Yes, gender effect. Women absorb omega 3 more so I guess that beneficial and detrimental effects, if any, are more obvious in women: Study: omega-3 uptake higher in women; we must account for gender in research, GOED VP says

Based on other studies (MR + RCT + animal models), DHA is the cause. Not EPA.

For what it’s worth I’m a male with above average EPA and DHA levels and I get depression from (EPA + DHA) but not from EPA-only (even at 2 g/day).

5 Likes
#73

Looking at this chart again, the most impressive thing is that there’s not a single subgroup that benefits from 1 g/day omega 3 (EPA + DHA) over 5 years. Even at this relatively low dose, supplementation was either detrimental or absolutely useless when it came to depressive symptoms.

3 Likes