There is mounting evidence that anti-oxidant supplements are cancer promoting (or perhaps more accurately, ‘deprotecting’). A lot of this data is in animal models, but there is also a bit of high quality controlled trial data. The best example is probably that vitamin E increases the risk of prostate cancer (https://jamanetwork.com/journals/jama/fullarticle/1104493).
Proponents will argue endlessly that it isn’t the right formulation of vitamin E and nitpick trial flaws, but taken as a whole the data is pretty clear. Eating AO rich foods is fine, but I would not take a supplement.
> Proponents will argue endlessly that it isn’t the right formulation of vitamin E and nitpick trial flaws […]
Vitamin E is not a single vitamin but is a group of eight different vitamins with different properties. Detrimental effects of large doses of α-tocopherol that was used in the study have long been implicated and cautioned about.
Need to take the literature as a whole. One can always postulate more complicated causal relationships and caveats, but in the end the question is should a random person take Vitamin E tomorrow? The answer is clearly no, for me.
A useful tool is Mendelian randomization. This looks for disease phenotypes that differ between people with different levels of vitamin E which is genetically determined. This is perfectly 'natural' vitamin E, not even a supplement.
It's a similar issue with Vitamin C, particularly intravenous vitamin C. Whenever a proper trial is done, it shows either not much, or harm (very significant harm recently: https://www.nejm.org/doi/full/10.1056/NEJMoa2200644).
The literature as a whole has not yielded conclusive results on either side of the fence yet. There are studies that have found benefecial effects of a supplement XYZ intake, and there are also studies that have found detrimental effects of the same XYZ supplement intake. We still have a poor understanding of how the body chemistry works and have an even poorer understanding of synergies or interrelationships between the intake of different «molecules». Moreover, how a particular chemical compound is going to affect to a random indiviual also largely depends on their overall health, age, sex, lifestyle and dietary habbits and a wide range of other compounding factors.
> … should a random person take Vitamin E tomorrow? The answer is clearly no, for me.
I am not advocating random people to start voraciously ingesting random things, however, the answer to such a question is far being clear. The short answer is «we don't know».
δ-tocopherol, for example, has no currently known biological role in the human body, therefore it will most likely undergo a safe passage in one's digestive tract and leave the body unharmed and without harming the body in the process. Most supplements will either give one a diarrhea or leave them constipated. Sometimes intermittently both. Lethal outcomes of the dietary supplement ingestion are not entirely unknown albeit are exceedingly rare.
There is a significant number of people out there, however, who start taking a supplement ABC expecting to develop supernatural abilities (e.g. grow a tail, acquire clairvoyance, learn the spell of casting a fireball or propel their endowment into unseen before highs) and, when nothing happens after 2-3 days, they become very disappointed or even disgruntled.
> This looks for disease phenotypes that differ between people with different levels of vitamin E which is genetically determined.
The Mendelian randomisation of the received/studied results is perfectly fine at the analytical stage of the study, however it is never a starting point of a new study. On the subject vitamin E specifically, the vast majority of studies use megadoses of α-tocopherol (400 IU or higher) that is not found anywhere in nature thus rendering study results inconclusive at best or useless at worst to the point that now most vitamin E studies can now be classified as pointless and junk, or as an easy way to obtain a research grant to later state that the study results are inconclusive and a further research is necessary.
The first study from your list, for example, focuses on measuring levels of vitamin E (only sketchily mentioning α-tocopherol levels) without quoting dietary sources (from food only or from food + supplemented?) across a random selection of samples from a blood test bank. The second study coyly mentions that «… diet‐derived antioxidants were used, which included vitamin E (α‐tocopherol and γ‐tocopherol) …» without specifying their respective sources (food or food + supplement). The third study is behind the Elsevier paywall and does not provide any meaningful details in the abstract.
One of the very first studies of vitamin E that has raised an alarm about potential harm of large doses of α-tocopherol was a Finnish one that followed a group of Finnish smokers over a succession of approximately 15 years AND supplemented them with large doses of α-tocopherol (not an abstract «vitamin E»!) throughout the entire study. The result of the study was an increased mortality across the study cohort which has been prescribed to misfortunes of the α-tocopherol intake. Yet, we still don't have a understanding of via which specific metabolical or chemical pathways α-tocopherol may influence the tumor development in the human body.
> This is perfectly 'natural' vitamin E, not even a supplement.
There is no such a thing as a «perfectly natural vitamin E» as «vitamin E» is a collective classification term for a cluster of up to 8 chemical compounds (α-, β-, γ-, δ-tocopherols and d-α-, d-β-, d-γ-, d-δ-tocotrienols) the concentration of which varies wildly across different food sources. For example, the red palm oil is unusually rich in tocotrienols whereas the soybean oil has them in nowhere near concentrations. α-tocopherol is also widely used as an officially approved food additive E306 (tocopherol) / E307 (α-tocopherol) which means that a unsuspecting broad population is exposed to an inadvertant larger intake of α-tocopherol as opposed to getting from natural source only.
You might think it’s a nitpick, but vitamin E is a fat soluble vitamin that usually coexists with other fat soluble vitamins. Taking any one of A,D,E,K without the others can cause problems.
That supports your concept of eating nutrient dense food. But I also think supplements can help a lot when done wisely (but most supplementation is probably not done wisely).
Proponents will argue endlessly that it isn’t the right formulation of vitamin E and nitpick trial flaws, but taken as a whole the data is pretty clear. Eating AO rich foods is fine, but I would not take a supplement.