Perhaps LifeExtension did a better job with there rebuttal here?
http://www.lef.org/f...Supplements.htm
That reply was sufficiently riddled with errors that when some of them were pointed out they took the post down. Edit: The reply in that link has its own problems, and isn't about the study under discussion.
I'm curious... About the problmes in the LEF rebuttal. Can you help? [...]
I feel the LEF rebuttal refuted the variables where most studies on this topic go astray. Can you point out where the LEF rebuttal goes wrong?
Well, let's start with the fact that their core argument is based on a notion that is not only complete speculation, but is actually completely opposite to the actual study details:
It’s a sad fact that most common diseases of aging are preventable, yet most people don’t engage in healthier lifestyle choices until after serious illness manifestation.
The classic example is an individual who never swallowed a single dietary supplement until they’re diagnosed with cancer. They then go from zero to low intakes of supplemental nutrients to swallowing 40 or more pills a day in what too often is a futile attempt to cure advanced stage disease.
In the Iowa Women’s Health Study,[(1) below] the authors admit they did not factor in the increased intake of dietary supplements that occur in response to the development of symptoms or diagnosis of serious disease. Stated differently: If a woman was diagnosed with stage 4 breast cancer and began ingesting 40 supplements daily, but died six months later, she would have been counted as being a heavy supplement user who died prematurely.
In other words, women who did not begin supplementing until after symptoms and/or disease manifested would have been classified from a statistical standpoint as being part of the group that ingested large quantities of supplements but died early nonetheless.
All emphasis in original. First, note that this is complete speculation: they have no
data to show that many women did this in this study, that many people do it in general, or that such people are, in fact, responsible for the excess mortality.
But it's worse than that: the rebuttal claims that "In the
Iowa Women’s Health Study, the authors
admit they did
not factor in the
increased intake of dietary supplements that occur in response to the development of symptoms or diagnosis of serious disease." But that's
absolutely untrue: the authors
explicitly state that they
did factor this phenomenon in, and it had
no impact on the results:An intermediate event, such as CVD or cancer, can induce a change in supplement use and confound the exposure-outcome association. In our data, the use of supplements was not modified by a pre-baseline diagnosis of CVD, diabetes mellitus, or cancer. Furthermore, intermediate cancer did not alter the supplement-taking pattern. ... In sensitivity analyses excluding women who had CVD or diabetes mellitus (n=5772) or cancer (n=3523) at baseline, the results were not materially changed. For example, for iron, the multivariable adjusted HR for total mortality was 1.13 (95% CI, 1.05-1.22). Parallel to the situation with total mortality rate, most supplements were unrelated to or showed higher cause-specific mortality rate in multivariable adjusted model version 2, although risk patterns varied across causes (Table 3). (1)
So:
they looked, and saw that women who developed one of these diseases were no more likely to start or stop taking supplements after diagnosis than the women in the rest of the study -- and just to be sure that any such effect was not biasing their results anyway, they did an analysis where they just
threw all such women out of the study, and it had no effect on the results: depending on the supplement,
either harm or no effect of supplementation in healthy women without pre-existing disease.So LEF's core argument is both false (they DID "factor in the
increased intake of dietary supplements that occur in response to the development of symptoms or diagnosis of serious disease") and had
no effect on the results. But now, let us (
for the sake of argument)
pretend that the authors had not done this, or that it had exerted a material effect on outcomes. LEF goes on to say,
This flaw by itself could render the overall findings of the Iowa Women’s Health Study meaningless because much of the lay public today mistakenly associates dietary supplements as something very important to initiate after serious disease appears.
Wha ...? If indeed "much of the lay public today mistakenly associates dietary supplements as something very important to initiate
after serious disease appears," and if indeed there were excess mortality caused by taking supplements under those conditions (especially if (
per LEF's false speculation) there was
not such excess mortality, or a positive countervailing
benefit, in women who had started taking them while healthy)), then this would make the results
even more meaningful, because those would be exactly the conditions under which the lay public would be most likely to take them!
(By the way: LEF has done plenty to encourage people with pre-existing disease to load up on supplements: their website and protocol book is loaded with advice to take a variety of supplements if you have various diseases, particularly cancer: see eg.
here,
here, and
here. So there's hypocrisy mixed in with their (conscious or unconscious) misrepresentation of the study analysis).
They then go on a long tear about confounding by hormone therapy:
In the Iowa Women’s Health Study, about twice as many women who took multivitamin/mineral supplements also took non-bioidentical hormone replacement therapy which is associated with early mortality. In numerical terms, 13.5% of the supplement users took hormone therapy at baseline compared to only 7.2% of non-supplement users. This ratio showing more supplement users taking hormone drugs persisted to the end of the observational period.
In technical terms, this is known as a ‘confounding factor’ because the increase in mortality caused by these dangerous hormone drugs would skew the results in a way that would show higher death rates in women taking these hormone drugs who also happened to be taking a multivitamin/mineral supplement.
Sure: this is absolutely an important consideration. So what do you
do with a potential confounding factor? You
perform an adjusted analysis to take the disproportionate use out of the picture. LEF (intentionally or due to failure to read the study) neglects to mention that the authors
adjusted for use of HRT:
in multivariable adjusted model, version 1, we additionally adjusted for educational level, place of residence, diabetes mellitus, high blood pressure, body mass index, waist to hip ratio, hormone replacement therapy, physical activity, and smoking status.(1)
The result? The magnitude of the increased mortality for multivitamin use went
up, not down: in the basic analysis (adjusted only for age and energy intake), multivitamin users
appeared to possibly be 2% more likely to die over the course of the next 19 years, but it was not statistically significant and might have been statistical noise or residual confounding; in multivariate-adjusted model 1, they were a statistically-significant
6% more likely to die.
That's the other thing. Notice that risk went
up after adjusting for "educational level, place of residence, diabetes mellitus, high blood pressure, body mass index, waist to hip ratio, hormone replacement therapy, physical activity, and smoking status".(1) The most likely reason is that
users of supplements were leading healthier lives all around at baseline and at followup
.At baseline, compared with nonusers, supplement users had a lower prevalence of diabetes mellitus, high blood pressure, and smoking status; a lower BMI and waist to hip ratio; and were less likely to live on a farm. Supplement users had a higher educational level, were more physically active, and were more likely to use estrogen replacement therapy (Table 1). Also, supplement users were more likely to have lower intake of energy, total fat, and monounsaturated fatty acids, saturated fatty acids and to have higher intake of protein, carbohydrates, polyunsaturated fatty acids, alcohol [don't scream "aha!": we're talking 3.9 vs. 3.6 grams of alcohol a day, or about a third of a standard drink -MR] , whole grain products, fruits, and vegetables. Similar patterns were seen in the 2004 questionnaire among 19 124 women (Table 1) and for individual supplements (eg, iron and calcium) (eTable 1; http://www.internmed.com).(1)
The whole "sick-user bias" thing is, in other words, not only (a) untrue and (b) irrelevant to the results — it's the
reverse of the case. The vitamin users were
healthy lifestyle people, who should if anything have had
lower mortality rates. And disease onset did not change their supplement use either way. Indeed, that has been the case in every study I've ever seen of suppelment users: uniformly, supplement users are
not unhealthy people who pop pills either to make up for a bad lifestyle or in response to disease diagnosis, but are (mild to extreme) health nuts, who exercise more, eat better diets, are thinner, better-educated, consume less saturated fat, etc etc (eg. (1-9)). Far from a "sick user" confounder, there is if anything a "healthy user" confounder: if anything, epidemiological studies
should find (as indeed many of them do, especially in unadjusted or poorly-adjusted analysis) that supplement users live
longer, due to all the other healthy stuff they're doing.
Indeed: when LEF argues that
The Iowa Women’s Health Study initially showed that women who supplemented with vitamins C, D, E and calcium had significantly lower risks of mortality.
The study authors then blended this positive data with the negative results of iron and copper use and other multivariable adjustments to conclude increased mortality associated with multivitamin/mineral supplementation.
These “multivariable adjustments” were at the discretion of the study authors and could have been skewed either way to show positive or negative outcomes. Based on the negative remarks expressed at the beginning of the report, there appeared to be author bias against the use of multivitamin/mineral supplements.
Now, first, you can't have it both ways. You can't bitch about the confounding by the fact that the supplement users were more likely to use HRT (and ignore/fail to mention that HRT was adjusted for), and
then bitch that the illusory protective effect of various supplments vanished after adjustmentfor the various ways that supplement users were different from nonusers (mostly, again, in a
healthy rather than unhealthy direction).
But second, these multivariable adjustments were
not "at the discretion of the study authors" (except in the trivial sense that no one was holding a gun to their heads to force them to do them): the authors looked, saw the various ways in which users were different from nonusers (younger, thinner, eating more veggies, exercising more, etc etc), and
adjusted for the observed confounders. If they had
not done so, LEF would have been justified in their (in the event, vapid) outrage over HRT confounding -- and other critics would have been justified to be outraged by failure to adjust for the healthier lifestyles of the users.
There's more bunk about the "European study," and embedded under "Too Much Vitamin A!!", "The Problem with Questionnaires," and "This Study Has Nothing to Do With Life Extension Members" (= "Move along; nothing to see here" = "Fnord!" = "Pay no attention to that man behind the curtain"), but I don't have all day ...
There's still compelling evidence though, that a relatively low-dose multivitamin/mineral supplement is beneficial. But you have to look close enough to see the important nuances and let neither the puffy LEF propaganda nor the recent anti-vitamin media hysteria blur your vision...
While interesting and non-crazy (and non-venal ...), I don't really find that terribly compelling evidence ... and while I don't have time for a microdissection of your analysis there, I will emphasize again that the "sick user" bias is really
not present in (either the great majority, or all) epidemiological studies of supplement users vs. nonusers: again, the normal (universal?) pattern is the
reverse.
References1: Mursu J, Robien K, Harnack LJ, Park K, Jacobs DR Jr. Dietary supplements and mortality rate in older women: the Iowa Women's Health Study. Arch Intern Med. 2011 Oct 10;171(18):1625-33. doi: 10.1001/archinternmed.2011.445. PubMed PMID: 21987192.
2: Dickinson A, Mackay D. Health habits and other characteristics of dietary supplement users: A review. Nutr J. 2014 Feb 6;13(1):14. [Epub ahead of print] PubMed PMID: 24499096.
3: Touvier M, Kesse E, Volatier JL, Clavel-Chapelon F, Boutron-Ruault MC. Dietary and cancer-related behaviors of vitamin/mineral dietary supplement users in a large cohort of French women. Eur J Nutr. 2006 Jun;45(4):205-14. Epub 2006 Jan 2. PubMed PMID: 16382374; PubMed Central PMCID: PMC1973945.
4: White E, Patterson RE, Kristal AR, Thornquist M, King I, Shattuck AL, Evans I, Satia-Abouta J, Littman AJ, Potter JD. VITamins And Lifestyle cohort study: study design and characteristics of supplement users. Am J Epidemiol. 2004 Jan 1;159(1):83-93. PubMed PMID: 14693663.
5: Ishihara J, Sobue T, Yamamoto S, Sasaki S, Tsugane S; JPHC Study Group. Demographics, lifestyles, health characteristics, and dietary intake among dietary supplement users in Japan. Int J Epidemiol. 2003 Aug;32(4):546-53. PubMed PMID: 12913027.
6: Kirk SF, Cade JE, Barrett JH, Conner M. Diet and lifestyle characteristics associated with dietary supplement use in women. Public Health Nutr. 1999 Mar;2(1):69-73. PubMed PMID: 10452734.
7: Lyle BJ, Mares-Perlman JA, Klein BE, Klein R, Greger JL. Supplement users differ from nonusers in demographic, lifestyle, dietary and health characteristics. J Nutr. 1998 Dec;128(12):2355-62. PubMed PMID: 9868181.
8: Patterson RE, Neuhouser ML, White E, Hunt JR, Kristal AR. Cancer-related behavior of vitamin supplement users. Cancer Epidemiol Biomarkers Prev. 1998 Jan;7(1):79-81. PubMed PMID: 9456246.
9: Houston DK, Johnson MA, Daniel TD, Poon LW. Health and dietary characteristics of supplement users in an elderly population. Int J Vitam Nutr Res. 1997;67(3):183-91. PubMed PMID: 9202979.
Edited by Michael, 18 February 2014 - 03:39 PM.