Supplemental Woe

Following on from the expose about McKeith and her crackpottery it is interesting that the news of late has been trumpeting the “dangers” of using vitamin and herbal supplements. Remember one of the main claims of the woo-ers supporting McKeith is that modern medicine kills and all these herbal supplements dont.

It seems (JAMA, vol 297, p842) that this is not the case. The report comes to the following conclusions:

Treatment with beta carotene, vitamin A, and vitamin E may increase mortality. The potential roles of vitamin C and selenium on mortality need further study.

The study found that people taking beta carotene supplements were at 7% greater risk of death than those who were not taking the supplements, and with Vit E it was 4% greater risk. More worryingly, Vitamin A supplementation appeared to increase the risk of mortality by 16%.

Now, there is a bit of a cautionary comment to go with this – this is a “meta analysis” study not a direct study, so there is the possibility that the people being given the supplements were at a greater risk of dying than the general public anyway, however one of the study group has commented “seventy percent of the participants were healthy.” (New Scientist)

Assuming this study is an accurate reflection, it is a nice slap in the face of those who push this woo in the place of retrovirals, immunisation and other “real” medicine 🙂

2 thoughts on “Supplemental Woe

  1. I take your point on this, and agree completely that lots of supplements are not harmeless.

    However, when I read this story, I found there were enough caveats to cast serious doubt on the message. It was a review of studies, some of which were of dying people. The report doesn’t make it clear whether the substances that were associated with a greater number of deaths were the ones that were tested on seriously ill people, which would be the first question. This seems most likely to me, as I’d imagine you’d need to study an enormous a huge number of people taking vitamins before you found enough deaths to count.

  2. I largely agree with you here, although the study did look at 232,606 people which is a pretty large grouping. The data was published in a total of 285 publications.

    While there are still questions, the data seems fairly strong here.

Comments are closed.