It seems to come as a shock to many people that so many "statistical" studies have been done that have found supplements and vitamins either useless ....or in some cases harmful! The latest, reported about ten days ago, was accompanied by media headlines to the effect that Vitamin E can actually trigger prostate cancer, for example. Many people's heads spun back and forth, since their physicians had told them earlier that Vitamin E in normalized doses, say of 400 iu per day, was beneficial. So was selenium.
Unmentioned in the media reports from the Cleveland Clinic research (supported by the National Institutes of Health) was that the abstract of the published paper clearly stated that the statistical results were not significant. Translated, that meant that they were aware that they had proven nothing! How can this be?
Earlier, in a blog last year, I'd explained how statistics can be used to obtain deceptive or erroneous results, e.g.
http://brane-space.blogspot.com/2010/07/more-damned-lies-with-statistics.html
I followed this by another blog to do with the specious statistics used in several statin studies (first reported by FORBES in 2008) wherein outcomes from a control and statin-taking group were compared per capita for heart attacks. In the control (placebo) group there occurred 2 heart attacks per capita, and in the statin-takers there occurred 1 heart attack per 100. The results were then touted as "Statins reduce risk of heart attacks by 33%".
Thus:
(2 - 1)/3 x 100% = 1/3 x 100% = 33%
Which was duly reported by the hobbled corporo-media and which has many people thinking: 'Wow! I need to get statins, because hey -- if there could be 1,000 likely to get 'em and 333 didn't because of taking statins, I wanna be in that number!"
Trouble is, there was no such magnitude of success, and the actual difference per capita for the trials was kept hidden, well, at least until FORBES exposed them.
It is entirely plausible that a similar "result" was obtained for the margins of difference in prostate cancer arising from taking vitamin E, but this time - because portions of the media were on to the ginnied stats, they decided to admit that in actual fact, the results were not significant.
Of course, there are tests that can bear this out! They are contingent on knowing actual sample sizes and then using the text to obtain the confidence level or level of significance.
For example, the well-known 'z-test' compares effects for two designated populations, P1 and P2, which have associated sample sizes N1 and N2. If the distinct outcomes for each population sample are then denoted by x and y, respectively, one can denote the separate populations in terms of the outcomes using:
P1 = x/N1 and P2 = y/N2
The z-statistic is then computed using:
z = (P1 - P2)/ [PQ {1/N1 - 1/N2}]^½
Where,
P = (x + y)/N1 + N2
is a pooled population and Q = (1 - P) (the variance of (P1 - P2) is the quantity in the denominator of the right -hand side in the expression for z)
Clearly, if a null hypothesis is applied such that no difference in outcomes is expected, one could have:
x/N1 = y/N2
or very close to that.
Now, here is a key element, we will know that z possesses a standard normal (e.g. Gaussian) distribution - see graph attached, if N1 and N2 are large. But, if the samples are too small, all bets are off and it is derelict to apply the z-test. For rejection of the null hypothesis (i.e. accepting a significant difference in outcomes) the obtained z-value cannot be less than 1.96 or the upper 5% of the standard normal distribution.
In effect, what the vitamin researchers are saying-admitting in their abstract, is that they didn't attain this threshold for the test they employed, which may not be the same as the z-test.
What was nefarious, however, and certainly not 'cricket' is that the study authors evidently continued collecting data for several more years - after the study ended - until they had met their pre-selected objective and found more men from their original group who had developed prostate cancer. Thereby they could insist according to their rewritten rules, that Vitamin E IS dangerous.
But this isn't allowable! Not scientifically or ethically! It would be as if I performed a published study on the incidence of a certain type of solar flare, and used ex post facto data - then years later went back and "adjusted" the study using concurrent data to alter the finding! The question is WHY would anyone resort to such tricks? Given the risk of discovery, however remote, the only plausible reason would be payoff to make the specious case.
In the case of vitamins- supplements, we know that ever since Durk Pearson's outstanding research on the benefits of vitamins in the 1970s, government and Big Pharma has been out to quash it. Why express surprise? We are talking of the Corporate State and it protects its interests! The polluted politicians know that if Big Pharma's profits are affected - especially by a bleed off from people spending on vitamins, supplements - they will lose campaign donations from the Big Pharma lobbies for the next election cycle. Duh!
Now, since most government agencies are already a case of the fox guarding the henhouse in terms of regulations, it makes sense that at any given time a particular agency that's been taken over by corporate power will inveigh against interests it deems a threat...say to PhRmA's enormous profits, and hence campaign donation capacity. Thus it is the politicians seek to stamp out even medical marijuana because they know it will dent PhRma's profits - just as the NIH now joins the fray to put a lid on vitamin purchases.
We cannot, CANNOT - have the little guys spending so much less on vitamins to stay healthy when they could spend a LOT more on prescription drugs by not taking them! I mean, think of it! If they can just get these little hardheads to stop taking a Vitamin C, planet sterol, aspirin combo each day - warning them off it - they will get their cholesterol and other numbers bad enough to warrant the prescription of costly statins! PROFITS! Wa-HOO!
Also, if Big Guv can squash the Medical MJ biz then what's not to like? The little cancer sufferers will have to then make use of a "synthetic" cannabinoid manufactured by Big PhRma at $10,000 a month - at least if they wish to ward off the worst effects of chemo!
Having taken over 14 vitamins and supplements for nearly 37 years I am not about to stop now, no matter what new "studies" proclaim. Also, as a person versed in statistics, I will have to see ironclad STRONG (not marginal) statistical evidence that a given outcome from use of a particular supplement is bad. Until that happens, all these specialists in negative vitamin research can go fart in the wind. It will be about that much difference to me in terms of whether they impact my thinking.
Let's also recall this mysterious 'white paper' that surfaced out of a think tank some 12 years ago, asserting that people - especially the elderly -need to be encouraged to do much more smoking, and gorging on fats, alcohol in order to "increase the death rates so we are able to pare down the entitlement costs".
Who wrote this crap? No one knows, but the theory is that it was one of the Right wing, capitalist think tanks. Maybe the same one that has been paying scientists (or offering to pay them) $10,000 for each article they publish against the theory of man-made climate change.
So...if this is possible...might there not also be a possibility that people are warned off vitamins to maybe increase their death rates, so massive "entitlement amounts" don't have to be paid out?Cynical and inquiring minds want to know.
Finally, I was gratified when in 1998, after years of skepticism over my vitamin taking, my then physician finally admitted to me openly during a physical exam that she'd been wrong, and she now concurred with my vitamin and mineral supplement regimen. As she put it:
"Well, I did the research ....all over the place, and I have to admit now that people simply can't get all the nutrients they need from food. It's all been so diluted, that a person would need to eat massive volumes just to get the minimum RDAs of one vitamin, far less what you're taking. So keep on keeping on..."
I just smiled and replied:
"I told you so!"
Unmentioned in the media reports from the Cleveland Clinic research (supported by the National Institutes of Health) was that the abstract of the published paper clearly stated that the statistical results were not significant. Translated, that meant that they were aware that they had proven nothing! How can this be?
Earlier, in a blog last year, I'd explained how statistics can be used to obtain deceptive or erroneous results, e.g.
http://brane-space.blogspot.com/2010/07/more-damned-lies-with-statistics.html
I followed this by another blog to do with the specious statistics used in several statin studies (first reported by FORBES in 2008) wherein outcomes from a control and statin-taking group were compared per capita for heart attacks. In the control (placebo) group there occurred 2 heart attacks per capita, and in the statin-takers there occurred 1 heart attack per 100. The results were then touted as "Statins reduce risk of heart attacks by 33%".
Thus:
(2 - 1)/3 x 100% = 1/3 x 100% = 33%
Which was duly reported by the hobbled corporo-media and which has many people thinking: 'Wow! I need to get statins, because hey -- if there could be 1,000 likely to get 'em and 333 didn't because of taking statins, I wanna be in that number!"
Trouble is, there was no such magnitude of success, and the actual difference per capita for the trials was kept hidden, well, at least until FORBES exposed them.
It is entirely plausible that a similar "result" was obtained for the margins of difference in prostate cancer arising from taking vitamin E, but this time - because portions of the media were on to the ginnied stats, they decided to admit that in actual fact, the results were not significant.
Of course, there are tests that can bear this out! They are contingent on knowing actual sample sizes and then using the text to obtain the confidence level or level of significance.
For example, the well-known 'z-test' compares effects for two designated populations, P1 and P2, which have associated sample sizes N1 and N2. If the distinct outcomes for each population sample are then denoted by x and y, respectively, one can denote the separate populations in terms of the outcomes using:
P1 = x/N1 and P2 = y/N2
The z-statistic is then computed using:
z = (P1 - P2)/ [PQ {1/N1 - 1/N2}]^½
Where,
P = (x + y)/N1 + N2
is a pooled population and Q = (1 - P) (the variance of (P1 - P2) is the quantity in the denominator of the right -hand side in the expression for z)
Clearly, if a null hypothesis is applied such that no difference in outcomes is expected, one could have:
x/N1 = y/N2
or very close to that.
Now, here is a key element, we will know that z possesses a standard normal (e.g. Gaussian) distribution - see graph attached, if N1 and N2 are large. But, if the samples are too small, all bets are off and it is derelict to apply the z-test. For rejection of the null hypothesis (i.e. accepting a significant difference in outcomes) the obtained z-value cannot be less than 1.96 or the upper 5% of the standard normal distribution.
In effect, what the vitamin researchers are saying-admitting in their abstract, is that they didn't attain this threshold for the test they employed, which may not be the same as the z-test.
What was nefarious, however, and certainly not 'cricket' is that the study authors evidently continued collecting data for several more years - after the study ended - until they had met their pre-selected objective and found more men from their original group who had developed prostate cancer. Thereby they could insist according to their rewritten rules, that Vitamin E IS dangerous.
But this isn't allowable! Not scientifically or ethically! It would be as if I performed a published study on the incidence of a certain type of solar flare, and used ex post facto data - then years later went back and "adjusted" the study using concurrent data to alter the finding! The question is WHY would anyone resort to such tricks? Given the risk of discovery, however remote, the only plausible reason would be payoff to make the specious case.
In the case of vitamins- supplements, we know that ever since Durk Pearson's outstanding research on the benefits of vitamins in the 1970s, government and Big Pharma has been out to quash it. Why express surprise? We are talking of the Corporate State and it protects its interests! The polluted politicians know that if Big Pharma's profits are affected - especially by a bleed off from people spending on vitamins, supplements - they will lose campaign donations from the Big Pharma lobbies for the next election cycle. Duh!
Now, since most government agencies are already a case of the fox guarding the henhouse in terms of regulations, it makes sense that at any given time a particular agency that's been taken over by corporate power will inveigh against interests it deems a threat...say to PhRmA's enormous profits, and hence campaign donation capacity. Thus it is the politicians seek to stamp out even medical marijuana because they know it will dent PhRma's profits - just as the NIH now joins the fray to put a lid on vitamin purchases.
We cannot, CANNOT - have the little guys spending so much less on vitamins to stay healthy when they could spend a LOT more on prescription drugs by not taking them! I mean, think of it! If they can just get these little hardheads to stop taking a Vitamin C, planet sterol, aspirin combo each day - warning them off it - they will get their cholesterol and other numbers bad enough to warrant the prescription of costly statins! PROFITS! Wa-HOO!
Also, if Big Guv can squash the Medical MJ biz then what's not to like? The little cancer sufferers will have to then make use of a "synthetic" cannabinoid manufactured by Big PhRma at $10,000 a month - at least if they wish to ward off the worst effects of chemo!
Having taken over 14 vitamins and supplements for nearly 37 years I am not about to stop now, no matter what new "studies" proclaim. Also, as a person versed in statistics, I will have to see ironclad STRONG (not marginal) statistical evidence that a given outcome from use of a particular supplement is bad. Until that happens, all these specialists in negative vitamin research can go fart in the wind. It will be about that much difference to me in terms of whether they impact my thinking.
Let's also recall this mysterious 'white paper' that surfaced out of a think tank some 12 years ago, asserting that people - especially the elderly -need to be encouraged to do much more smoking, and gorging on fats, alcohol in order to "increase the death rates so we are able to pare down the entitlement costs".
Who wrote this crap? No one knows, but the theory is that it was one of the Right wing, capitalist think tanks. Maybe the same one that has been paying scientists (or offering to pay them) $10,000 for each article they publish against the theory of man-made climate change.
So...if this is possible...might there not also be a possibility that people are warned off vitamins to maybe increase their death rates, so massive "entitlement amounts" don't have to be paid out?Cynical and inquiring minds want to know.
Finally, I was gratified when in 1998, after years of skepticism over my vitamin taking, my then physician finally admitted to me openly during a physical exam that she'd been wrong, and she now concurred with my vitamin and mineral supplement regimen. As she put it:
"Well, I did the research ....all over the place, and I have to admit now that people simply can't get all the nutrients they need from food. It's all been so diluted, that a person would need to eat massive volumes just to get the minimum RDAs of one vitamin, far less what you're taking. So keep on keeping on..."
I just smiled and replied:
"I told you so!"
No comments:
Post a Comment