The references were a couple of sentences before. Miller et al. 2004, "Meta-Analysis: High-Dosage Vitamin E Supplementation May Increase All-Cause Mortality", "Randomized Trial of Vitamin A and Vitamin E Supplementation for Retinitis Pigmentosa." Though I see the formatting made it look like 1 reference instead of 2, and the link was broken.

Also, this is something you can easily google if you question it.

People familiar with vitamin studies would know right away that 200,000IU of vitamin A and 5,000IU of vitamin E per day are both extremely high, and not in the same category as "vitamin supplementation".

Sure. "Known to be toxic" was probably too glib a way to put it, Phil should provide a cite and I shouldn't have repeated it uncritically. But even if this was just someone's hypothesis without much experimental evidence behind it: the concept of vitamin poisoning isn't a new one. There are publicized daily intake tolerable upper levels for lots of vitamins. Hypervitaminosis A, E. I don't know what kind of evidence backs these claims up but presumably people in the field are aware of this kind of thing. I'm not saying "These are the toxicity levels. This is why the meta analysis is wrong." I'm saying "People are hypothesizing vitamin toxicity at certain levels. Why the hell would you run an analysis that couldn't even in principle take that into account?"