Examining the Evidence
A study with 4 inescapable limitations
Four key problems limit the effectiveness of this study.
1. Inappropriate population
Seventy percent of the women in this study were between the ages of 60 and 79 with a mean age of 63. This fact disqualified the WHI’s first report as a primary prevention study of cardiovascular disease, and it has a major nullifying impact on this study as well.
Only a small percentage of older postmenopausal women have vasomotor symptoms; in this study just 12% noted them as “moderate to severe.” Women with severe vasomotor symptoms were dissuaded from joining the study due to inability to take placebo. In her editorial, D. Grady comments that, “Among the 12% of women who did report moderate-to-severe vasomotor symptoms at baseline, the symptoms were unlikely to be very bothersome, since the women were willing to be randomly assigned to placebo.”1 Hence, this is not an appropriate population from which to draw conclusions about quality of life issues.
2. Discontinuation rate was 42%
This was an intent-to-treat study and almost half of those in the study group discontinued therapy. This is certainly not unexpected when women are arbitrarily placed on a single estrogen-progestin combination therapy without regard to individualizing treatment–especially when you consider that 88% were without vasomotor symptoms at baseline. The breast tenderness, bloating, bleeding, headaches, and mood changes from a 1-size-fits-all regimen would be enough to make most women discontinue treatment if their clinicians were unable to adjust their therapy.
Subjects who stopped therapy remained in the treatment arm for determination of quality of life results. The authors admit that “it is possible that differences were not significant at 3 years because of…poorer adherence to assigned therapy.”
3. The conjugated equine estrogen/medroxyprogesterone acetate combination in this study does not represent all HRT formulations
The definition of what constitutes HRT is vastly different today than it was a mere 20 years ago. Thus, it is impossible and misleading to extrapolate the WHI results to the many different options of estrogens, progestogens and delivery systems presently available in the US.
Numerous studies have shown estrogen-associated increases in quality of life. Progestogens, especially medroxyprogesterone acetate–the most potent synthetic progestin we have–can attenuate these estrogen benefits by down-regulation of the estrogen receptor. This is a process we seek in the endometrium, but want to avoid in brain, bone, vascular tree, genitalia, and skin. Better progestogen choices now available, such as micronized progesterone, norethindrone acetate, and norgestimate, are less potent and far better tolerated in combination with the many estrogen options.
4. Quality of evaluation tools
This study attempts to evaluate quality of life using various medical scales–each designed to assess a specific function, but none developed to actually measure quality of life. The most primitive scale, utilized to evaluate “sexual satisfaction,” consisted of just 1 question with 4 choices: very unsatisfied, a little unsatisfied, somewhat satisfied, or very satisfied. Other researchers have utilized vehicles with 40 questions on a 10-point scale in studies of sexuality, and the academic sexual societies are constantly trying to evolve more sophisticated tools to evaluate this complex concept. One question simply cannot assess sexual satisfaction.
Individualization of therapy has been, and should continue to be, the guiding principle in helping patients decide whether or not to begin HRT, and ultimately which combination best fits their needs. This unsatisfying study uses the wrong population, continuation, combination, and evaluation and fails to consider the variations in genetic complement of estrogen receptors. No single therapy is appropriate for all women.
1. Grady D. Postmenopausal hormones–therapy for symptoms only. N Engl J Med. Posted March 17, 2003. Because of its potential therapeutic implications, this article was published early at: www.nejm.org. It appears in the May 8th issue.