Medscape today website reported on Oct. 10, 2011 “Vitamin Supplements Associated With Increased Risk for Death”
The study is entitled "Dietary Supplements and Mortality Rate in Older Women" - Arch Intern Med. 2011;171(18):1625-1633
When I read the details of this study, I had to ask myself – doesn’t anybody teach research methods at Universities anymore? When I was a graduate student one of our professors prescribed a little book that covered about 3 lectures in a statistical methods course.
The book title was “How to Lie with Statistics” and this study made me think about it again. The study was described as a “Cohort Study” which implies longitudinal observations, but it would be more correct to call it an observational, questionnaire study because there was no attempt to control selection of subjects (except for age and gender) and no controlled experimental interventions, just observation.
They called it “New findings” from the Iowa Women’s Health Study. This is a study, started in 1986 and completed in 2004 where 41,836 postmenopausal women aged 55-69 completed 3 questionnaires regarding supplement use (1986, 1997 and 2004) — the study actually involved 5 questionnaires - including 1989 and 1992 but only 3 asked about vitamin supplements.
They only used data from 38,772 participants – no explanation about why. Even though the last data collection was in 2004, they included deaths up to 2008. There were 15,594 or as they emphasized - 40% but didn’t specify how many died after 2004. Naturally with this age group you would expect some deaths due to various causes. The major causes of death up to 2004 were 4,900 from cancer and 5,900 from cardiovascular disease which leaves 4,794 from other causes including accidents, murders and suicides.
The study concluded that multivitamins and other selected vitamins showed and increased HAZARD RATIO for “total mortality risk” when compared with corresponding non-use. The highest risk ratio was for supplemental copper at 1.45 (See more in #8 below), Folic Acid at 1.15, supplemental iron at 1.10 (likely the most prescribed vitamin by physicians), Vitamin B6 at 1.10, zinc at 1.08, magnesium at 1.08 and Multi-vitamins at 1.06.
The authors and the apparently biased media don’t bother to discuss other results like the improved hazard ratios for B complex, Vitamin C, and Vitamin D. The headlines and the study seemed to emphasize an alleged increase in mortality risk for multi-vitamins which even with their distorted statistical analysis was only a 2% increase above non-users whereas iron showed a 9% risk and is likely one of the most prescribed nutrients of all time. Then there is copper with a 31% increase in mortality risk. Again this must be prescribed because I certainly don’t know of large numbers of people running out to buy supplemental copper and I know it is carefully controlled as an ingredient in multi-vitamin.
Flaws in Study Design
- Self-reported questionnaires have a wide range of reliability. People lie!
- There was no indication of the health status of the participants upon entering the study.
- There were no controls over the participants health status when they started using supplements. The reported usage increased throughout the duration of the study and supplement users had a higher educational level, were more physically active, and were more likely to use estrogen replacement therapy.
- No indication of why they started supplementing.
- There were no differences reported that indicated dosage, frequency nor compliance. (once/day, twice/day, every other day, weekly, when I remember??).
- There was no differentiation between synthetic or natural source supplements.
- There were no physiological explanations offered as to why some supplements or any supplements for that matter would contribute to increased risk of death.
- Warped statistics were very evident upon examination of data available in the tables published in the study. An example is the distortion over the copper supplement. The sample size of the women taking copper in this study was a grand total of: 24 people by the end of the study. There were 30 in 1986, 57 in 1997 and only 24 in the last session of questions and yet they claim a 31% increase in mortality rate.
- Effects of vitamin supplementation was not the primary purpose of the study. It was one of those “Oh we have some data here, lets see what is correlates with”. The Iowa Women’s study was designed to
- Determine if the distribution of body fat (waist/hip) predicts incidence of chronic diseases, with the primary endpoints being total mortality, and incident cancers of the breast, endometrium, and ovaries, and
- Determine to what degree diet and other lifestyle factors influence risk of chronic disease.
- The Council For Responsible Nutrition states “Further, the authors show their own bias with this statement: ‘We recommend that they [dietary supplements] be used with strong medically-based cause, such as symptomatic nutrient deficiency…’ which basically means these researchers would rather wait till we all get scurvy before acknowledging any need for supplemental nutrients.
The Council emphasizes that dietary supplements are commonly taken to help prevent chronic disease.
- A further bias of the publication is shown when it invited a guest commentary by well known critic of supplements, Dr. Goran Bjelakovic, MD, from the Centre for Clinical Intervention Research, Cochrane Hepato-Biliary Group. His opinions on supplements and their potential role in good health is already well-known, and whose own work has been the subject of controversy in scientific circles. I have written about him before. Read more
- I have to say that the Main Stream Media are somewhat complicit in the distribution of this distorted misinformation. Please realize that the Archives of Internal Medicine is owned by the Journal of the American Medical Association and both journals rely heavily on Drug company support for their funds that maintain their ability to publish. In the same thought pattern, the drug companies are among largest users of print advertising in main stream media. You be the judge.
- Rarely do you see front page attention for supplement studies that demonstrated positive health benefits like the Landmark study that I have written about before.
Of course, the biggest hoax in all this is that there has never been a single confirmed case of a vitamin killing anyone, and yet FDA-approved pharmaceuticals kill at least 100,000 Americans every year!
Prescription drugs are actually the fourth-leading cause of death in this country, according to a study in the Journal of the American Medical Association (remember the Viox scandal and the 60,000 heart attacks)