Supplements
Articles
Flawed Iowa Women's Health Study Used to Discredit Supplements: Don't Believe It!
by Charles Poliquin
10/17/2011 1:50:03 PM
Last week a research study associating regular supplement use with increased mortality received an enormous amount of media attention. The Iowa Women's Health Study published in the October edition of the Archives of Internal Medicine was fundamentally flawed, making the study practically useless. The media is exacerbating the problem by latching on to the limited and flawed conclusions made by the study’s authors, and the already inadequate results are being portrayed in confusing ways! Let’s see what the study really says and what we can learn from it.
What Are the Details of the Study?
The study was an 18-year observational study of older white women living in Iowa. The purpose was to test if taking supplements helps reduce mortality rate. The study collected self-reported diet and supplement information from over 38,000 women using a variety of questionnaires in 1986, 1997, and 2004. The supplements asked about were multivitamins, vitamins A, D, E, and D, beta-carotene, the various B vitamins, iron, calcium, copper, magnesium, selenium, and zinc. Self-reported data on body composition, general diet, and disease was gathered.
Researchers attempted to isolate specific factors that resulted in greater death rate by “controlling” for weight, diet, disease, and smoking status among other things. Their control methods are questionable, in part because they were using complicated statistical calculations to control for numerous interrelated factors. Plus, they didn’t control for aspects that likely contributed to death rates, such as hormone replacement therapy or prescription medication use.
Initially, results showed that women who supplemented with the vitamins C, D, E, and calcium had significantly lower rates of death over the study period. Iron and copper supplementation were linked to greater risk of death. Then, researchers compiled all supplement use data to conclude that women who use one or more of the fifteen supplements had a slightly increased rate of death (over 15,000 women died over the 18 years). At the start of the study the average age was 61 and at the end it was 82.
Most concerning was the prevalence of iron supplementation, and taking iron was most strongly associated with death. In comparison, taking calcium was associated with less risk of death over the study period.
Isn’t Age 82 Older than the Average Life Expectancy for Women?
That’s right. What is strange about the study is that it is measuring how many women die over an 18-year period, ending at age 82, which is older than the current average life expectancy for women in the U.S. According to a 2006 congressional report, the average life expectancy for women was 80 as of 2003. This means that over 50 percent of the population in the Iowa study lived longer than the average life expectancy.
It’s interesting but problematic that researchers analyzed the relation between mortality rates and supplement intake in individuals who died during what was likely a fairly normal age range for expected mortality. The fact that over 50 percent of the studied population lived longer than the average life expectancy is not addressed in the study.
How Did Researchers Control for Physical Activity and Other Factors?
Researchers “controlled” for factors such as age, diet, weight, physical activity level, and smoking status, meaning that they used statistical methods to eliminate those factors from skewing the results. The concern is that the questions used to gather data were very generalized. For example, physical activity was characterized as “moderate activity less than a few times per month, once a week, or twice a week.” There was no information regarding type, intensity, or duration of physical activity.
If you’re reading this, you probably know the critical importance of physical activity on health, and that type, intensity, and frequency vary drastically and can have significant effects in promoting health and longevity. Although exercising twice a week is better than nothing, it’s certainly not ideal for preventing disease and prolonging life.
A complicating issue is that the data was self-reported, calling its value into question when so many different confounding factors are present. Researchers appear to agree that their controls may not have worked because, although they performed extensive statistical adjustment for cofactors, they write that it is possible “residual confounding remained.”
Self-reported questionnaires are notoriously unreliably in correctly gathering information. When participants self-report behavior, they typically overestimate and evaluate it more positively than behavior actually is. Additionally, the questionnaires were not validated. Self-reported questionnaires need to be validated in order to ensure that they assess what they are intending to. Validation is a scientific process that ensures accuracy of a questionnaire and is necessary in order to even begin to take conclusions made from a self-reported survey seriously.
How Was Diet Controlled For? Did they Consider Organic vs. Conventional Foods?
A 127-question dietary questionnaire was used, which may appear to be exhaustive, but it doesn’t account for food quality, whether food was organic or conventional, or if participants regularly ate foods with artificial additives, such as those found in soda. Food additives, pesticides, and growth hormones used on conventional meat could skew results significantly because they are all known to have a profound effect on health, and they can hinder absorption of vital nutrients. If participants were eating large amounts of conventional and artificial foods, they may have completely negated any positive effects from supplementation, but we don’t know since this information was not surveyed.
We do know that the women who used supplements increased their average calorie intake by over 150 calories during the 18-year study period, which could be a complicating factor since higher calorie diets have been shown to shorten life expectancy. The women who didn’t report any supplement use also increased calorie intake over the study period, but only by an average of 60 calories a day.
Average macronutrient intake was very similar between supplement users and non users and didn’t change over the 8-year study period. Diets consisted of an average of 18 percent protein, 34 percent fat, and 48 percent carbohydrates. This is a fairly low-carb, high-fat diet that could be improved by increasing protein intake, but overall it is probably preferable to one that is higher in carbohydrates.
Wouldn’t Medications Skew the Results if they Weren’t Controlled For?
Yes! Medications weren’t surveyed, although researchers did calculate the number of women on hormone replacement therapy for menopause. Naturally, prescription medications can cause countless negative side effects, but they may also help prolong life, meaning they can’t be ignored when considering longevity. Failing to control for prescription drug use is another reason the study is seriously flawed, making it impossible to identify an association between supplement use and mortality.
Researchers found that 13.5 percent of supplement users were on hormone drugs at the start of the study. Women who didn’t take supplements were less likely to take hormones with only 7.2 percent on drugs in 1986. These numbers dropped by 3 to 4 percentage points by the end of the study in both groups.
It’s significant that almost twice as many of the supplement users took hormone drugs, which were later shown to have significant negative health effects, including increased cancer and death rates. For example, one of the most common hormone replacement drugs of the 1980s and ‘90s was Premarin, which was later linked to increased breast cancer risk. Indeed, a 15-year Women’s Health Initiative study launched in 1991 collected a variety of health data, including information about hormone drug use. Part way through the study, the Data and Safety Monitoring Board recommended that women discontinue hormone replacement therapy drugs such as Premarin because “hazard ratios” were too high in the form of increased breast and lung cancer risks.
The double rate of hormone drug use in the women who took supplements is a major confounding factor that wasn’t controlled for—the slightly higher rate of mortality could actually be due to hormone drug use. But, since researchers didn’t even attempt to control for hormone therapy, the study’s integrity is seriously compromised.
Iron Use Increased Mortality. Why Were the Women Even Taking Iron?
Researchers found a 3.9 percent increase in mortality associated with iron supplementation in women free of cardiovascular disease, diabetes, or cancer. Also, researchers found iron use increased as participants aged, and that progressively low doses were associated with increased death risk as the women got older.
Why were these women taking so much iron, and why was supplementation frequency increasing as the women aged? We don’t know because this issue wasn’t addressed in the study discussion, another concern with how the researchers reported and explained the data. Adequate iron is important for optimal health, and iron deficiency can cause anemia and negatively affect brain function, among other things.
But, iron deficiency is mainly a concern for children and women of reproductive age due to blood loss from menstruation. Pregnancy may increase dietary iron needs as well. In men and postmenopausal women, excessive blood loss such as regular blood donation, or having a diet extremely low in foods containing iron (vegetarians are at risk) could result in low iron. Individuals with chronic disease may be at greater risk of low iron. For example, one study of diabetic men found that 24 percent had anemia, and low iron was linked with low testosterone.
But, the women in the study weren’t pregnant, of reproductive age, and they didn’t have a chronic disease. We don’t know if they were vegetarians or regularly donated blood, but even if a few women did, it wouldn’t explain the large use of iron supplements. Unfortunately, researchers don’t address the women’s elevated iron use aside from noting evidence of the harmful effect of high iron levels. Indeed, men with high iron have been found to have increased rates of cardiovascular disease. This is likely because high iron levels results in the production of oxidants or free radicals (the opposite of antioxidants that support health), which cause oxidative stress and lead to heart disease.
What Other Factors May Explain the Results?
A major concern is that popular multivitamin brands of the study period tended to have excess levels of vitamin A, a nutrient that has been linked to increased bone loss, and fracture incidence. And there’s evidence that in the elderly, a hip fracture is a precursor to health decline that can increase death rate. One study of Swedish women found that vitamin A intake was negatively associated with bone mineral density, and for every one mg increase in daily vitamin A consumption, risk of hip fracture increased by 68 percent. This is because vitamin A is known to counteract the ability of vitamin D and calcium to build bone density, and excess vitamin A leads to bone loss.
Vitamin A supplementation is necessary for individuals who are deficient, such as children or individuals who are malnourished, but deficiency is not a common problem in American adults. It is likely the women in the Iowa study who took vitamin A or a multivitamin had elevated vitamin A levels, which probably countered the numerous beneficial effects of vitamin D. Take note that vitamin D researchers have suggested that if the world population doubled its vitamin D intake, it would decrease death rates by 10 percent!
How Did Researchers Account for Quality in Supplements Used by the Women?
They didn’t! Researchers note they did not have any details of the type or quality of the nutritional supplements that were taken by the women. There is an enormous variation in the quality and potency of supplements available on the market. Cheaper supplements are just that, cheaper in quality as well as cost. In light of the strange iron supplementation and the high rates of hormone therapy, it’s unlikely the women in the study were well informed about the importance of high-quality supplements, making it a good bet that they generally opted for lower quality, cheaper supplements, although we don’t know for sure.
Supplement quality can be compromised at many stages of the production process and the most common problems are contamination and low potency. When companies want to cut cost they use cheaper raw materials, which result in a lower quality product. For example, in the case of magnesium, magnesium oxide is a cheap, small molecule that isn’t absorbed well by the body, whereas magnesium glycinate and citrate are large molecules that are harder to package but much more easily absorbed. This type of magnesium is effective at raising blood serum levels, but it is also more pricey to produce. Low quality supplements like magnesium oxide aren’t absorbed by the body. Rather, they pass right through and can cause negative side effects—magnesium oxide can have a laxative effect.
Additionally, the best delivery option for optimal absorption of a vitamin or mineral varies. For example, vitamin D is best delivered in liquid gel capsules, meaning if you’re getting yours in a hard tablet, you’re probably not getting much vitamin D. Plus, when vitamins and minerals are extracted from their natural form and made into a supplement, they may become unabsorbable by the body.
Then there’s the issue of potency. If supplement manufacturers don’t test raw materials and the final product for potency, the product that makes it to the shelf may be utterly useless. Additionally, contamination is a major concern. Raw materials may contain toxins, solvents, heavy metals, harmful bacteria, yeast, mold, pesticides, or dioxins. If you’re taking a cheap supplement with heavy metals and toxins in it, your good intentions are being completely thwarted!
Think about your favorite steak dinner—if you start with a poorly nourished, possibly growth hormone-filled, sick animal, you’re going to get a low quality, poor tasting, contaminated steak. Does it make you want to become a vegetarian? Probably not, and no need to. Rather, pick the organically raised animal and the high-quality, most potent, most absorbable supplement!
Speaking of diet, it makes me wonder if the women who were taking supplements were eating well rounded diets. Taking a low quality supplement that provides no benefit for the body may in fact exacerbate nutrient deficiencies if those nutrients are avoided in the diet because an individual thinks they are getting the nutrients elsewhere. For example, if you’re supplementing with fish oil daily, but not eating fish due to high mercury levels, you may be missing out if your fish oil is not potent or is filled with toxins. Even though the Iowa study didn’t test for fish oil use, I’d like to mention it here because it is a perfect example of how low-quality supplements can hinder the best of intentions for good health.
High-quality omega-3s are well known for having numerous health benefits, including preventing heart disease and improving cognitive function. Take note that because low-quality omega-3s are derived from fish, they have been found to contain numerous toxins and contaminants. Indeed, in 2009 a lawsuit was filed against some of the largest supplement manufacturers (Rite Aid, GNC, CVS, NOW Health Group, Solgar, and Twinlab among others) because independent lab tests found PCB toxins in the fish oil sold by these companies.
PCBs are used in coolants and are banned in the U.S. due to harmful health effects including cancer and neurological disorders in fetuses. A decision hasn’t been made in the lawsuit, but if you want to learn more about what I found out, check out my article
Fish Oils and PCBs: The Cost of Cutting Corners.
Are There Any Studies That Show Supplement Use Decreases Mortality?
A report published in the online July issue of the European Journal of Nutrition found that vitamin and supplement use significantly reduced cancer and mortality rates in a group of men. This study also used self-reported questionnaires and has multiple factors that could confound results. Nonetheless, it should have been mentioned and addressed by the researchers in the Iowa study because it is common ethical practice to include a literature review that considers contradictory and supporting evidence. The Iowa study researchers failed to address it, likely because it produced contradictory results from their study.
Another concern is that researchers attempted to single out specific nutrients and relate them to rate of death. This is an ineffective method because most diseases are highly complex and take years—probably a good portion of a lifetime—to develop. Genetic predisposition is another factor, as is the complex nature of foods and how they affect the body.
For example, foods such as almonds or red raspberries that are nutritionally rich aren’t just high in vitamin E or vitamin C, but also contain other nutrients and chemicals that support health(for example, red raspberries are the best source of antioxidants called ellagitannins and contain a decent amount of fiber and protein). This may seem like a no brainer for the Poliquin readership, but the best strategy is to eat a very wide variety of organic foods and use high-quality supplements to support detoxification, while treating nutrient deficiencies.
What Can I Take Away From This Flawed Study?
Take away the knowledge that you can’t blindly trust research studies. Peer-reviewed research can generally guide us in making the best health choices, but a reasonable amount of skepticism is best. Also, be extremely cautious regarding claims made by the media about your health. In this age of business-driven news, outrageous headlines that misrepresent well-researched studies are the norm, and the media takes flawed studies like the Iowa Women’s Health Study and magnifies the useless conclusions with language that will grab your attention. Find a source you trust for your health news, question what you don’t, and be extremely skeptical with claims made by mainstream media.
References:
Mursu, J., Robien, K., et al. Dietary Supplements and Mortality Rate in Older Women: The Iowa Women’s Health Study. Archives of Internal Medicine. October 2011. 171(18), 1625-1633.
Li, K., Kaaks, R., et al. Vitamin/Mineral Supplementation and Cancer, Cardiovascular, and All-Cause Mortality in a German Prospective Cohort. European Journal of Nutrition. July 2001. Published Ahead of Print.
Johansson, S., Melhus, H. Vitamin A antagonizes Calcium Response to Vitamin D in Man. Journal of Bone and Mineral Research. October 2011. 16(10), 1899-1905.
Melhus, H., Michaelsson, K., et al. Excessive Dietary Intake of Vitamin A is Associated with Reduced Bone Mineral Density and Increased Risk for Hip Fracture. Annals of Internal Medicine. November 1998. 129(10), 770-778.
Coad, J., Conlon, C. Iron Deficiency in Women: assessment, Causes, and Consequences. Current Opinion in Clinical Nutrition and Metabolism Care. September 2011. Published Ahead of Print.
Goddard, A., James, M., et al. Guidelines for the Management of Iron Deficiency Anemia. Gut. October 2011. 60(10), 1309-1316.
Kocjan, T., Prelevic, G. Hormone Replacement Therapy Update: Who Would We Be Prescribing this to Now? Current Opinion in Obstetrics and Gynecology. December 2003. 15(6), 459-466.
Slatore, C., Chien, J., et al. Lung Cancer and Hormone Replacement Therapy: association in the Vitamins and Lifestyle Study. Journal of Clinical Oncology. March 2010. 28(9), 1540-546.
Katalinic, A., Rawal, R., et al. Decline in Breast Cancer Incidence After Decrease in Utilization of Hormone Replacement Therapy. Breast Cancer Research Treatment. 2008. 107, 427-430.
Cushman, M., Kuller, L., et al. Estrogen Plus Progestin and Risk of Venous Thrombosis. Journal of the American Medical Association. October 2004. 292(13), 1573-1580.
Anderson, G., Judd, J., et al. Effects of Estrogen Plus Progestin on Gynecologic Cancers and Associated Diagnostic Procedures: the Women’s Health Initiative Randomized Trial. Journal of the American Medical Association. October 2003. 290(13), 1739-1748.
Grossmann, M., Panagiotopolous, S., et al. Low Testosterone and Anemia in Men with Type 2 Diabetes. Clinical Endocrinology. April 2009. 10(4), 547-553.
??Carreo, J., Barany, P., et al. Testosterone Deficiency is a Cause of Anemia and Reduced Responsiveness to Erythropoiesis-Stimulating Agents in Men with Chronic Kidney Disease. Nephrology Dialysis Transplant. May 2011. Published Ahead of Print.
Willcox, D., Willcox, B., et al. The Okinawan Diet: Health Implications of a Low-Calorie, Nutrient-Dense, Antioxidant-Rich Dietary Pattern Low in Glycemic Load. Journal of the American College of Nutrition. August 2009. 28(Suppl 500S).
Shrestha, Laura. Life Expectancy in the U.S. CRS Report for Congress. August 2006. http://aging.senate.gov/crs/aging1.pdf
Copyright ©
Back to top