There are two types of middle aged women – those who do and those who don’t. Those women who feel younger than their age, and those who feel older. Those women who embrace life with energy, and those who slog through the day as if might be their last. Those women who are pleasant to be around, and those you want to avoid at all cost. Those who accept hot flashes and other physiologic changes as a part of aging, and those who don’t. We’re talking about hormone replacement therapy (HRT) of course. Over the next several articles we will delve into the topic of HRT for women.
There may not be another area of medicine, especially in women’s health, that is more confusing than hormone replacement therapy – and unnecessarily so.
Misconceptions: Why Some Women Do and Some Don’t
In the upcoming articles we will address the more common misconceptions of hormone replacement therapy to better answer why some women do and some don’t take HRT. And, we will start by addressing a study called the Women’s Health Initiative (WHI) that has dragged women’s health back to the dark ages rather than catapult it into the future.
We begin that journey of misconceptions, misperceptions, and misguided conclusions regarding HRT by first focusing on the statistical mumbo jumbo that clouds the picture as it relates to the WHI and HRT. We will focus on the WHI because that is the study that gave HRT in women a black-eye from which it has not fully recovered. Plus, WHI represents the worse portrayal of HRT. But, even then the findings aren’t not nearly as bad as originally made to be and still stated to be.
Keep in mind there are numerous studies before and since the WHI, a bulk of which show that the hormones including the synthetic or non-bioidentical used in WHI are safe.
Statistical Significance versus Clinical Relevance
Let’s start with a brief discussion on statistical significance and clinical relevance. Findings in a study are said to be statistically significant, or not. Findings that are statistically significant are not to have occurred by chance alone. In other words, the results are real – the findings accurately representing what they claim, while clinical relevance begs the question “do the findings really matter?”
Findings can be statistically significant – be real – yet have no real meaning in the real world. That is they are not clinically relevant. So much of the results found in the WHI are of borderline statistical significance. But, more importantly they are of questionable clinical relevance. But, we tend to focus on statistical significance and struggle sometimes to take a step back and look at the bigger picture – the real relevance of something.
For instance, are you concerned if a study finds a statistically significant additional risk of breast cancer with a particular intervention even though the additional risk is only 0.45% when the intervention most likely will make you feel better and lower the risks of other problems? Maybe – but maybe not depending on your view and individual circumstances. If you have a strong family history of breast cancer such a small additional risk may be enough to concern you. For many though it probably does not.
Women’s Health Initiative
The Women’s Health Initiative was designed to be a two armed primary prevention trial comparing estrogen alone (in the form of Premarin) and combined use of estrogen plus progesterone in the form of Premarin and medroxyprogesterone acetate (Provera) with placebo. The use of estrogen plus progesterone was in the form of a combination pill called Prempro. Outcomes looked at included the development of heart disease, breast cancer, stroke, pulmonary embolism, endometrial cancer, hip fracture, and death due to other causes.
The results of the first arm – combined estrogen (Premarin)/progesterone (Provera) in form Prempro – were reported in 2002. Here is how the media reported those findings.
- 26% increase in breast cancer related to Prempro
- 29% increase in heart attacks related to Prempro
- 41% increase in strokes related to Prempro
Anyone who read that would have been concerned and any woman on HRT would have been scared. But, these percentages are misleading. If 2 out of 100 people get a heart attack. And, that number goes up to 3 out of a 100 people when placed on a drug then that represents a 50% increase in heart attacks even though the additional risk is actually only 1% (one more heart attack per hundred). Let’s take a closer look at the WHI numbers. Here’s how the study reported the numbers.
Outcomes were reported as cases per 10,000. The World Health Organization has designated/defined 10 cases or less per 10,000 as a rare event. Below in table form are the outcomes related to the use of Provera found in the WHI Study.
WHI 2002 Findings: Prempro Versus Placebo (events per 10,000 patients)
|Outcome||Prempro versus Placebo|
|Breast cancer||8 more cases|
|Heart attack||7 more cases|
|Pulmonary embolism||8 more cases|
|Stroke||8 more cases|
|Colon cancer||6 fewer cases|
|Hip fracture||5 fewer cases|
Again, these are cases per 10,000 patients. So that’s an 0.08% and 0.07% increase in breast cancer and heart attacks respectively – statistically significant but hardly clinically relevant. The overall additional risk in the Prempro arm was 19 events per 10,000 person-years or 0.19% additional risk.
Are you willing to accept the overall risk of 0.19% that some adverse effect might happen to you to go on a drug that will eliminate your hot flashes, improve your mood and cognitive function, lower your risk of hip fracture and colon cancer, make your skin look better, give you better energy, help you control your weight plus more? Plus, that is zero point nineteen percent using a drug foreign to your body. Do we have reason to believe that hormones identical to what your body makes would fare worse?
Here are the results from the Premarin arm alone published in 2004.
WHI 2004 Findings: Premarin Versus Placebo (events per 10,000 patients)
|Outcome||Premarin versus Placebo|
|Breast cancer||7 fewer cases|
|Heart attack||No effect|
|Pulmonary embolism||12 more cases|
|Stroke||6 more cases|
|Colon cancer||No effect|
|Hip fracture||6 fewer cases|
The overall additional risk in the Premarin arm of the 2004 WHI Study was 2 events per 10,000 person-years or 0.02% additional risk. Interestingly, the decrease in breast cancer in the Premarin arm did not garner the same media attention as the increase in breast cancer in the Prempro arm even though the magnitude in change was similar.
When we look at the differences in heart attacks we can conclude that it is not estrogen (Premarin) at fault, but rather the progestin (Provera) that was in Prempro. Provera acts very differently in the body than the natural progesterone the body actually makes. We’ll discuss this more in an upcoming article in this series.
The Numbers Behind the Numbers
I like to say it’s the numbers behind the numbers that matter most. The lifetime risk of a woman getting breast cancer is 12.4% (one in 8 women). There are estimated to be 231,840 new cases (American Cancer Society) of breast cancer in 2015. In the Prempro arm 166 out 8,506 women or 1.95% who received Prempro got breast cancer while 124 out of 8,102 or 1.5% who received placebo.
So in that arm there was an additional risk for breast cancer of 0.45%. Are you willing to accept that additional risk if Prempro controlled your hot flashes, improved your mood, gave you better energy, and alleviated the other symptoms that go along with being post-menopausal? When presented that way most women would say ‘yes’. The risk of anesthesia or any surgical procedure is higher than 0.45%. And, the risk of something bad happening to you as drive to and from work probaby exceeds 0.45%.
Consider that based on the rate that breast cancer grows it is likely that all women who developed breast cancer during the 5.2 mean years of follow up in WHI 2002 likely had malignant cells in their breasts at entry into the study – at least that is the conclusion of R. Don Gambrell, MD. In fact, some women in the study only received Prempro for 2-3 years – not enough time to produce a breast cancer large enough to be detectable. In other words, they most likely had occult breast cancer before entering the study. (Read the full study)
Also, there were fewer hip fractures resulting from osteoporosis in both arms of the study. Osteoporosis affects 40% of women. In 2010 there were 258,000 hip fractures (CDC) with 193,500 occurring in women. One out of 5 people who suffer a hip fracture die within one year. READ that AGAIN. Not many women die within a year of being diagnosed with breast cancer. This risk reduction and death reduction related to osteoporosis and hip fracture received barely a mention by the media and has been overlooked by physicians.
Now you may look at the numbers from WHI and say “yeah, but overall there is still an additional risk from HRT”. True in the WHI there was increased risk, but they are of borderline statistical significance and of questionable clinical relevance. Also, the study made no attempt to quantify the symptomatic improvement that women on HRT obtained – the main reason women seek in taking HRT. Also, there have been studies before and since WHI using the same hormone formulations that have shown HRT to be safe.
Perhaps more importantly the WHI used two synthetic hormones. Premarin is actually several horse estrogens, at least 10 of which the female does not make, and two of which are known be carcinogenic in women (but not carcinogenic to horses from which it is derived). Provera is a progestin which is not the same as the progesterone the female body produces.
So if the risk from synthetic hormones in the worst case scenario study (WHI) is barely statistically significant, what are the risks related to bioidentical hormones – the very same hormones the human female body makes and which have been “clinically” tested in real life in real human beings for at least 6,000 years of recorded history?
Also, ponder this. If hormones are so dangerous why don’t we see more of these adverse events when women are younger and have higher hormone levels? Why is it that the incidence of anxiety/depression, heart disease, obesity, high blood pressure, diabetes, cancer, and strokes begin to rise as hormone levels begin to decline below critical threshold levels? Pure coincidence?
And, this takes us to the next article where we discuss differences between bioidentical (natural) and non-bioidentical hormones (synthetic).