Posted on 02/28/2007 2:45:16 AM PST by XR7
Thanks so much!
I have been in the pharma business for a long time (and worked in pharma nutrition for ten years also) and I take one Centrum Silver per week, an extra few grams of folic acid every week but 3 tablets/day of vitamin D (GNC's "Vitamin D 700") as cholecalciferol. FWIW.
Fair enough. I'm not signing on to the full-on tinfoil hat brigade on this. I was just a little suspicious of this study.
You're absolutely correct about MSM/Chrondroitin/Glucosamine for arthritis. I forgot about that. It works.
3mg of C a day?
That doesn't even cover the USDA/FDA minimums, which is what, a minuscule 68mg, just enough to keep scurvy at bay.
Are you sure you don't mean 3 grams?
As I'm a smoker, I take around 5~6g a day, and I can feel the difference. The doc also put my dad on CoQ-10 and suggested I do the same. So I'm taking about 200mg a day.
I've also read that among mammals, there are only three that do not produce their own C and require it from diet/supplements. Humans, Gorillas, Hamsters, and there might be one other, can't recall.
And IIRC, blood serum levels of other mammals that do produce their own C extrapolated to human weight/mass, would result in serum levels in the 10's of grams.
Realized that I wrote 3 mg. inst. of (3000 mg. or) 3 g., but figured I'd wait until someone else noticed it ;-)
LOL! And I was only thinking that meta-analysis might just be compounding exponentially the inadequacies of the earlier studies! You're right though -- it's the wave of the future!
Regards, Ivan
"Meta-analysis might just be compounding exponentially the inadequacies of the earlier studies!"
Yes, I had not thought of this, but you are right. That's 2 things conceptually wrong with Meta-analysis. Meta-analysis as described so far is not a valid technique.
I once took a course in advanced statistics and experimental design. I believe our instructor would have laughed at any of us who had proposed anything like this for a real science study.
Physicians generally pooh-pooh the value of vitamins, but this is because you can get vitamins into your body without the aid of physicians. If the only way it could be done was for a team of eight surgeons to implant a special $263,000 trapdoor in your head, physicians would say vitamins were the best thing since luxury German automobiles.
I always trust Dave Barry! ;-)
. . . or, as James Taranto of Best of the Web Today put it: To 105%?
Growing old isn't so bad when you consider the alternative.
I'd like some evidence that the people who did this study ever took an undergraduate statistics course.
The figures from the best quality trials on selenium however showed that it might reduce death risk by 10 per cent, either on its own or in combination with other supplements, but this was not found to be statistically significant.
Are they saying that a 4% increase in death stats is significant but a 10% decrease in death is insignificant? Makes you wonder what their bias was...
Hi Coleus....that's interesting.
Someone told me to take hawthorn for my blood pressure. I tried it, but perhaps I wasn't taking the correct dosage, so I resorted to a medication my physician gave me. I was living dangerously with hypertension for too long.
Bjelakovic et al. published similar articles in The Lancet (2004), and in Alimentary Pharmacology & Therapeutics (2006). The flaw in all three studies is the same. They used multiple linear regression, a technique that averages in people who took moderate doses of vitamins along with people who took highly toxic doses of vitamins.
The average dosage for vitamins A and E in their new study were at levels already known to be toxic; the maximum dosages were about 20 times the recommended levels. Of course there was some mortality - just as a study of people who took 1 to 100 tablets for a headache, and pretended the dose-response was linear, would show that "aspirin increases mortality".
People frequently cite the many numbers and measures of statistical significance in Bjelakovic's studies as evidence of the validity of their results. Statistical significance means nothing if the test being performed is nonsensical.
Dr. Bjelakovic has not responded to my request to see his data in order to run a multiple nonlinear regression test on it. It is fairly certain, from prior studies, that such a test would show that vitamins A and E have beneficial effects up to dosage levels of about 5,000 and 300 IU/day, respectively.
Appropriate dosage levels depend on age, sex, and weight, as well as personal genetics and medical condition - so be careful out there.
More details on how to correctly interpret the study data are on my livejournal page, shagbark.livejournal.com .
- Dr. Phil Goetz
Thanks.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.