Research Bias in the Women’s Health Initiative (WHI) Study
Vinay Prasad often writes about bias in research. In my field, I also see bias in research studies. I have always considered the research on Premarin biased. I don’t like the 60 years of Wyeth studies “demonstrating” that Premarin was good for the heart, vessels, breasts and brain. I know it was good for hot flushes, osteoporosis, bladders and vaginas. The Federal Drug Administration’s (FDA’s) endorsing the Wyeth studies for sixty years is reprehensible.
For the better part of 60 years, Wyeth watched low-density lipoprotein (LDL) cholesterol as an endpoint (the primary outcome that is being measured by a clinical trial) rather than using the cause of death or age as an endpoint. The erroneous big assumption was that lower LDLs would cause better vascular health overall and therefore delay death from strokes and heart attacks in particular, the number one cause of death for postmenopausal women. This belief was so common and so firmly promoted by the medical community that some doctors refused to care for postmenopausal women who did not want to take Premarin.
While the Women’s Health Initiative (WHI) study was not perfect from my perspective, it drove home two important points. The first finding demonstrated that contrary to what Wyeth and the FDA had told us for sixty years, Premarin and Provera were neither effective nor safe. These meds actually increased heart attacks, breast cancer and dementia so much that the WHI study was stopped prematurely.
Choosing the correct endpoint is a detail that both Wyeth and the FDA should have determined six decades earlier. The fatal flaw resulted in the premature deaths of not thousands but millions of women. If I were involved with the premature or avoidable death of a postmenopausal women I would be sued and lose my license. Big pharma and the FDA are immune from fines and criminal charges. This too-friendly relationship between Big Pharma and big government should not have happened and should certainly not still be happening today.
Believe it or not, settling the “safe and effective” issue of HRT is what the WHI study got right. But immediately when I read the study in 2002, with its three groups or participants (one given Premarin alone, a second given Premarin and Provera, and a third given Premarin with natural progesterone) I asked myself where was the fourth arm, the one with estradiol and progesterone?
I very much dislike the term hormone replacement therapy (HRT) because its too generic. For example, the term car can refer to many different kinds of cars. What’s wrong with saying Buick or Mercedes? So, what’s wrong with saying estradiol, or conjugated equine estrogen (Premarin). The problem is that 98 percent of the population, including some doctors, don’t see any difference between the various kinds of estrogens. When we hear the term vaginal estrogen we are supposed to divine by some magical means that the vaginal estrogen being referred to is Premarin vaginal cream made from conjugated equine estrogen, a euphemistic term for horse urine.
Where is the fourth arm, the arm with natural estrogen (estradiol) and real progesterone? Certainly, we know that before menopause there is estradiol, the real estrogen made by ovaries. We also know that heart attacks in women are much fewer before menopause and much more after it. This correlation appears to equal causation. To my knowledge there have been no studies of the combination of estradiol and real progesterone.
In my country doc way, I think that the WHI study was almost as wrong as the Wyeth-FDA studies and approval. The differences are two. The Wyeth FDA studies actively promoted Premarin, now known to be harmful for some indications in some women, whereas the WHI study damage was passive, withholding estradiol because it has been considered to be an estrogen lumped into the same category as Premarin. Of course, the study demonstrating the benefits of estradiol hasn’t been done, so it is hard to promote estradiol use, something I consider a real problem.
Not doing an estradiol arm in the WHI was for me an obvious error, an error that should have obvious from the start of the study. However, a second less obvious error was the WHI failure to control for age and time of menopause in subjects regarding the initiation of HRT. Recent studies indicate HRT needs to begin at the time of menopause. These results have led to what has been called the “timing hypothesis.” In other words, beginning hormone replacement therapy soon after menopause provides better results than starting it later after a sustained time duration between menopause and starting hormone replacement.
With new studies, HRT is experiencing a comeback. But in the discussion of estrogen replacement, the term used is often simply estrogen. It is often unclear whether the estrogen under discussion is bioidentical estradiol or some other form of estrogen.
I believe there needs to be a study of estradiol in HRT. Any discussion of HRT should be clear about what form of estrogen is being referred to. Women cannot make decisions about their choices in HRT in cooperation with their doctor without knowing that there are various kinds of hormonal estrogen and be made aware that estradiol is the hormone their body is no longer producing.
As a physician having lived with the WHI study from two decades ago, researchers are consumed by the question “What happens to patients who are treated with HRT?” They should instead by asking “What happens to those patients who are not treated with estradiol, the real female hormone their body is no longer producing?