One of the serious problems we have to confront in medical research is the fact that much medical research and investigation has been focused predominantly on white men.
Over the past several years, as recognition of this problem has increased, more attention has been paid to including women and people of color in various clinical research programs.
The implications are significant, because we are now have less confidence that something discovered to work in white males or white men and women applies to other ethnic groups the same way.
One recent study that has received wide attention and has had significant impact is the Women’s Health Initiative (WHI). This study, which reported several years ago on the significant medical risks of hormone replacement therapy in women, resulted in a significant decrease in the use of hormone pills in post-menopausal women.
One key observation from the WHI study was that women who took combination hormone replacement therapy with estrogen and progestin had a higher incidence of breast cancer, especially if they had taken it recently over a long duration. Subsequent reports indicated that in women who had a prior hysterectomy and only took estrogen as their hormone replacement did not have an increase in the risk of breast cancer.
But most of the women participating in that study were white and only about 10% were women of color.
Since breast cancer occurs less frequently in black women, but has a worse outcome, can we assume that the effects of hormones are the same in black women?
A study reported today in the Archives of Internal Medicine asks that question, and provides some insight that is interesting but may not be definitive.
The article in question reports the results of a several year study called the Black Women’s Health Study. The study began in 1995 with the intention of following a large number of black women aged 21-69 over the years with questionnaires to determine their health status. 64,500 women participated in this research program.
Among the questions asked were those related to the use of hormone replacement therapy, menopausal status, and the development of breast cancer.
23,304 women were age 40 or over either at the start of the program, or turned 40 during the years 1995-2003 and were analyzed as part of this report.
The authors reported that 615 women developed breast cancer during the study. Of these, 364 never used hormones; 40 had used hormones in the past; and 210 used hormones recently. Of the recent users, the majority (134) used estrogen alone, while 67 used combination therapy.
The bottom line, according to the information presented in the article, was that long term use of hormone replacement therapy increased the risk of breast cancer by 58% in women who had used hormone therapy for more than 10 years, compared to those who had never used hormones.
Estrogen alone increased the risk of developing breast cancer by 41% in women who used the medication for 10 years or more. Combination therapy with estrogen and progesterone increased the risk of developing breast cancer by 45% in women who used the medicines for 5 years or more.
Perhaps most interesting was the observation that among women using hormone replacements, the drugs had a greater impact on the risk of developing breast cancer in thinner women using hormone replacements than on more obese women who used the drugs.
At first glance, this study would appear to say that if you use hormone replacement therapy, and you are black, you have an increased risk of developing breast cancer no matter which type of hormone therapy you used, so long as the use was recent, and 5 or 10 years more in duration.
But that is not consistent with the larger WHI trial noted above. In that study, the increased risk of getting breast cancer if you used the combination of hormones was 26% after 5 years, and, if you took only estrogens (these women all had hysterectomies) your risk of breast cancer was actually decreased.
What we have here is different information from two studies in two basically different ethnic populations. So what explains the differences? Is it due to racial differences, or some other factor?
The answer, in my opinion, is that the two studies are not comparable and the study reported above may not be the final answer to this question.
First, the studies are fundamentally constructed differently. The Black Women’s Health Study collected information over several years, but did not assign women to groups where one group took hormone treatments and the others did not. The WHI on the other hand was a randomized controlled trial, and although I have had questions regarding some aspects of the program, I cannot fault the fundamental study design and benefits of a randomized clinical trial. This research approach is always more preferable to an observational study, even one that collects data in a forward-looking manner.
But there is a more fundamental question for me in reviewing the currently reported study, and that is in what we call “statistical significance.”
It is difficult to explain “statistical significance,” so bear with me as I try.
In simplest terms, when you compare data from one group to data from another group you measure what is called “statistical significance.” That is a measure of how “solid” the differences are between the two groups, and whether or not there is a possibility that the difference may be due to chance.
If the groups are large, and the differences substantial, then the “statistical significance” is going to indicate that the differences are real. But if the numbers are small, and the differences close to each other, then the statistics will show that there is a possibility that the differences could be due to chance alone, and cannot be relied on with a great deal of confidence.
In the Black Women’s Health Study, the number of women in each of the two groups who developed breast cancer (those who took hormone replacements and those who did not) are small.
In addition, the increased risk ratios are not large for the most part, and when you examine what we call the “confidence intervals” we see they vary widely. That means the numbers for the most part do not have a great deal of statistical difference (or separation), which would allow me to place great confidence in their findings.
In other words, there is too great a possibility that the numbers are due to chance alone.
Because of that, unfortunately, I cannot necessarily agree with the authors that their findings prove the point that any hormone therapy will necessarily increase the risk of breast cancer in black women.
That does NOT mean that it is safe to conclude the alternative, namely that all hormone therapy is safe for black women. It is simply that this particular study at this particular time doesn’t allow me to draw the conclusion that there are differences in risks of hormone therapies between black and white women.
I would anticipate that if this study continues for many years, the quality of the data will improve considerably. The vast majority of the women in this study were in their 40’s, which is not the primary age when women develop breast cancer. Follow these women for another 20 years, and the data may indeed become more informative.
As to the question of whether leaner women (defined as a BMI less than 25) who used hormones had a greater risk of developing breast cancer than more obese women, the data in that case are a bit stronger, particularly for women who used estrogen for more than 10 years.
Why that is the case is uncertain. We do know that in post-menopausal women there is a greater risk of breast cancer in obese women than women of normal weight. This is thought due to the fact that fat cells increase the amount of estrogen circulating in the blood.
Why additional estrogen in hormone pills would increase breast cancer risk only in lean women is uncertain, and is deserving of further study.
So, bottom line, we still do not have an answer as to whether black women who use hormone medications differ in the risk of developing breast cancer compared to white women who use hormone replacement therapy.
Hopefully, the authors will continue their study and update their results periodically. For many women of color, there is too little information available that helps them understand whether or not they have risks different from other ethnic groups in our population.
As scientists and others continue to debate the differing effects of biology, race, and access to care, these types of questions are of too great a practical significance to ignore. We need solid, dependable information to address some very real and practical questions.