Special Feature

Death by a Thousand Indignities: The Saga of HRT

By Sarah L. Berga, MD

This week there were 2 articles in the New England Journal of Medicine that explored the link between postmenopausal hormone use and heart disease. The first article was the final analysis of the Prempro® arm of the WHI. The other was an interventional trial that asked if hormone use would reverse or halt the progression of established coronary artery atherosclerosis. The results of these studies can only be put into perspective by logic. Sadly, when it comes to the topic of postmenopausal hormone use, logic seems to have been replaced by hysteria and polarization. Appropriately, new data are reconciled with previous data, with the result that hypothesis and concepts are refined. In contrast, the tendency has been to view the newer data as supplanting previous data and as completely invalidating prior concepts and hypotheses. Some moderation would seem to be in order.

Let’s step back from the fray and examine one of the overarching, but often unspoken, issues fueling the current controversies.

It was expected that reproductive hormone use by menopausal women would prevent or reverse most, if not all, of the debilitating conditions of aging. Why? This expectation reveals an element of wishful thinking and the fallacies of directly extrapolating from bench to bedside. There was also an understandable collusion among doctors who wanted to help, women who wanted to do all that they could to stay well and attractive as they aged, and the pharmaceutical industry that had hormonal preparations to sell. There were some data to support the recommendations, but dose-finding studies were not done for each tissue expected to benefit from these actions. It was assumed that the dose needed for bone or for quelling hot flashes was the dose needed for other tissues. To be honest, there was molecular, cellular, and physiological evidence that had accrued over many years to suggest that most, if not all, tissues in the body were targets of hormones. There were reasons to think that the judicious use of reproductive hormones after menopause might retard aging or at least some of its clinical consequences. But we never were clear as to whether postmenopausal hormone use was intended for prophylaxis or whether it might be used as an intervention for established disease processes. The first study by Manson et al1 reports the results of using HRT as prophylaxis and the latter by Hodis et al2 using it as an intervention.

Given the results of the Prempro® arm of the WHI, we now realize that hormone therapy administered in a nonphysiological manner in a dose that is good for bone and hot flashes does not necessarily prevent cardiovascular disease and may have risks. We should not be completely surprised. What if we gave other hormones in the same manner and expected all who got them to benefit? Hormones are just another of the many agents being promoted for amelioration of the aging process in general, and cardiovascular disease in particular, that have fallen victim to the chasm between molecular promise and clinical application. Further, failure to demonstrate the desired effect may have as much to do with the manner in which the agent was given or other features of the study design as with the biological rationale for and conceptual foundation behind the study. The question to ask about any study is whether or not it provided a judicious test of the fundamental biological hypothesis. Typically, the answer for any one study is that it provided a partial answer. For instance, a clinical study may show us how not to give an agent, but it may not invalidate the biological concept behind giving the agent. There may be room for refinement or complete revision of the clinical approach. All study designs have limitations, and an honest investigator is typically quite circumspect about the increment in knowledge that each study provides. For some reason, we are now being asked to abandon circumspection in favor of the "bandwagon du jour." We were asked to believe first that reproductive hormone use by postmenopausal women was "all good." Now we are being asked to believe that it is "all bad." Neither of these polarized positions seems reasonable in light of the extant knowledge base.

I wonder what would happen if we were to give a set of key manuscripts to some unsuspecting college students with no axe to grind and ask them to interpret the evidence accrued to date? I would contend that there is no way to grasp the overarching concepts fueling these investigations and to put the current evidence in perspective without resorting to analogy (in addition to a formal, regimented analysis). For instance, by way of analogy, what if we gave all older women an insulin modifier such as metformin at a standard dose in perpetuity because we know that aging is associated with a decline in insulin sensitivity? What if we gave all women a standard dose of thyroxine because we know that metabolism declines with aging and we wanted to give it a boost? What if we gave standard doses of growth hormone by injection to all women older than 50 regardless of body size and composition? Would all benefit equally? Would all benefit? The reason to use standard doses is ease of administration, clinical practicality, and the need to manufacture standard preparations. The con is that too much or too little of any agent may lead to untoward or null clinical results. It had been assumed, but not proven, that the window for benefit was wide open for reproductive hormones, but this assumption is unlikely to be true if one considers what is known about dosing and clinical impact for other hormones like thyroxine, cortisol, and insulin. The nuances of the clinical practice of HRT have not been tested in any clinical trials of reproductive hormones. Physiology has never been used as a guide for study design for postmenopausal administration of reproductive hormones, while physiology is the guide in most other contexts.

There are even greater questions about study design when we extrapolate from what is known about the cellular and physiological effects of a hormone to an interventional disease model. For instance, what if we asked if reproductive hormone use could reverse kyphosis in older women because we know that reproductive hormones promote bone accretion in adolescence? I think that we would agree that this expectation is unrealistic, that kyphosis cannot be reversed, and that the inability of reproductive hormones to restore vertebral integrity does not invalidate the idea that reproductive hormones have important tropic effects for bone. The point that I am trying to make is that there are good reasons to suspect that hormones are important to health and that they play a role in aging. There are reasons to suspect that altering the endogenous hormonal milieu that accompanies aging might ameliorate to some extent certain age-related conditions. However, we would be foolish to think that this would be as simple as giving the same hormonal preparation in a one-size-fits-all dose regardless of age and health risks and years since menopause. Would that health promotion and chemoprevention of aging were so simple!

To my way of thinking, we still have more questions than answers. We need to find some middle ground and not get too polarized. We need to acknowledge that the newer studies have revealed limitations in the standard clinical approach of giving all postmenopausal women the same dose of the same product.

Questions that remain include:

1. Are all estrogens the same (ie, does the type, route, or dose matter)?

2. If dose matters, what is the appropriate range?

3. Are all progestins the same?

4. Does progestin use attenuate or augment estrogen’s effects?

5. Is the effect of progestin tissue specific?

6. Is there a SERM that confers a more favorable risk:benefit ratio than any of the traditional estrogens?

7. Will some form of HRT synergistically impact the benefits of lifestyle variables (ie, will those with a healthy lifestyle benefit from some form of HRT while those who smoke, lack exercise, or practice poor nutrition will not)?

Sadly, the search for a reductionistic panacea to the vagaries of aging has led again to disappointing results. The only certainty is that the search will continue, if for no other reason than most of us want to do something to retard the aging process and maintain health as best we can. This quest is an inherent attribute of a healthy individual and not an objective to be lightly dismissed.


1. Manson JE, et al. N Engl J Med. 2003;349:523-534.

2. Hodis HN, et al. N Engl J Med. 2003;349:535-545.