Main Menu              Health Library            Supplement on Heart Disease

USE of FEMALE HORMONES after MENOPAUSE

Abstract:  Life Ahead provides a powerful and now unique method for valuing the time related and complex effects on life of use of Post-Menopausal Female Hormones.   Hormones  increase the risk of breast cancer and female organ cancer, but reduce risk of colorectal cancer.  The substantial increase in risk of endometrial (uterine) cancer produced by estrogen can be largely removed by use of projesterone.  Hormones also increase slightly the risk of stroke.  Hormones increase the risk of heart disease during a first few years of use probably due to increased blood clotting and thrombosis.  But they also reduce the rate of atherosclerosis that produces a steadily lowering of heart disease risk that increases with duration of hormone use.  After about 5 years of estrogen and projesterone use the reduction in risk of heart disease usually offsets the increased risk of cancer and hormones produce a net benefit to risk of death.  Lifetime use of a this combination usually should  produce a modest benefit to life.   The Life Ahead analysis challenges some of the implications derived from recent WHI study that that used Premarin that produced high thrombosis, and was terminated at much too short a time to adequately value benefits from the longer term use of Female Hormones.  Research also shows that risk of hormones will not be higher for those surviving from breast cancer that for those cancer-free.  But use of such hormones must remain a decision of individual women and their doctors.  More on the risk of heart disease from female hormones is provided as a supplement.                          

Background:  Millions of women long have taken estrogen and estrogen-progesterone combinations after their menopause to avoid some annoying and painful troubles, and to gain cited advantages such as reduced risk of osteoporosis, heart disease, colon cancer, and some other health problems. But research has shown that hormones also increase risk of stroke, breast and female organ cancer.

The problem of judging the benefit or potential harm of using hormones is complex. First, use of post menopausal hormones for sufficient time has been shown by most research to reduce the risk of coronary heart disease, but increases the risk of cancer. Second, the balancing of these risks is difficult because women coming out of menopause can have differing base risks of each of the involved diseases from their family history and other risk factors.  Third, the research on these diseases and especially on heart disease has recently become confusing and a subject of intense controversy because of the publication of a new study that will be discussed following.

Life Ahead provides what may be a unique method for resolving a most probable contribution of female hormone use to health and life based on their effect on cardiovascular diseases and cancer. The risks to each individual woman user for each involved disease are quantified by Life Ahead at at their entry to the taking of hormones from a multiplicity of all major risk factors including details of diet and exercise. Second, the program computes a specific year-by-year life table of additional disease progression for each user for each involved disease, from time of starting hormones to any year of stopping them or for continuing their use to probable year of death.  Third, the risk of hormone use on each disease at each age is quantified by a global analysis of all major research found. The Life Ahead computation then produces results conveniently expressed both in Well-Days of healthy life.  If desired the likely cumulative risk of each involved disease from present to any age of life can be computed from time of starting hormones after menopause.

Heart Disease:  More than 40 different research study comparisons have confirmed that post-menopausal female hormones used over sufficient time reduce risk of heart disease substantially.  This benefit usually more than offsets the risks incurred from increases in risk of cancer.  Thus hormones can not only bring comfort to many women, and reduce risk of a variety of other problems such as osteoporosis, but reduce risk of major disease and death.    This conclusion was challenged during 2002 by publication of a study that suggested that hormones could increase - and not decrease risk of heart disease. If true, this would change the picture regarding the risk of using hormones long range. This new study called WHI was given enormous hype by the media and upset thousands of doctors and millions of women that have been taking hormones.

An extensive review of all of the population study research on female hormones found in the Life Ahead project showed that the significance of this new study was incorrectly overstated by the media.  Estrogen has two opposing effects on heart disease.  It- or  perhaps just some formulations of it - increase blood clotting, and this increases heart disease risk somewhat during a first 2-3 years of its use. But estrogen also reduces the atherosclerosis that over the longer range steadily reduces risk of the disease by a little more with each year of its use.  Thus after about 3-4 years hormones usually start to produce a net reduction in risk that steadily increases during future years, and eventually reduced risk of coronary disease by about 40-50%.  Hormone types that do not produce thrombosis may also reduce risk of heart disease in earlier years of use.

The new WHI study heralded as the "largest and most important study yet" actually was far smaller than several other studies including an important one on the nurses that showed hormones to have large benefits.  The Life Ahead analysis of the risk for use of hormones on heart disease is provided separately for those interested.  A table showing the extensive results of actual research data on hormones and heart disease is included with this separate analysis. This analysis favors conclusions from most past research that usual type hormones used in most past research probably will reduce coronary risk over the long term and provide some if very modest long term benefit to health of most women taking them.  The specific hormone premarin used in this study appeared to produce large amounts of thrombosis and this may have been responsible for much of the differing than past results obtained in this study.

Stroke:  The results of key research relating post menopausal use of hormones to stroke are summarized in Table 2 appended.  The risk ratios show consistent but modestly higher risks of stroke for hormone users. Omitting the high risk for the unusually high dose of estrogen in study H4 an average risk of 1.2 is computed for all results, with a 1.15 average for estrogen and 1.31 for estrogen + projesterone.  The clinical studies produced results similar to those from the other studies within their accuracy of estimate. Although the difference between the estrogen and estrogen + projesterone risks is of limited significance, it was felt best to include this because of the good comparison in large study S4.  Life Ahead thus assigns single factor risks for hormone use on stroke as 1.15 for estrogen and 1.25 for estrogen + projesterone  These values or even the average value of 1.20 for hormone use are well within the research error margins of risk ratios on stroke.

The results for risk of stroke from hormones are puzzling.  Most risk factors affect ischemic stroke and coronary disease similarly. Both heart disease and stroke are related to atherosclerosis, and atherosclerosis develops gradually over time. Yet there is no indication that the ‘estrogen-times-years’  factor applies for stroke risk.  This was clear both from the global analysis of results and from the specific review of stroke risk vs. years of exposure in study S4.  The short term risk that applies to heart disease – as perhaps thrombosis -  may be more important to risk of stroke and, may mask the contribution from changes in atherosclerosis.  It seems evident that a much better biochemical understanding of the chemistry of the hormone factors that produce disease is needed.

Breast Cancer:  Estrogen and particularly that produced during the pre-menopausal years is responsible for much of the breast cancer experienced by women. Breast cancer risk increases about 12.5% each year during this period of life, and declines to an increase of only 3% per year after the menopause.  Each year that first period or menarche is delayed reduces risk of breast cancer by nearly 6%.  Each year that menopause is earlier than usual reduces risk by 3.5%.  An earlier removal of ovaries produces the same reduction in risk as that for an earlier menopause. These studies confirm that risk of breast cancer multiplies with years of estrogen exposure, or produces a risk as a function of estrogen times years present. Life Ahead computes the added risk of breast cancer for each year from first period to menopause as discussed separately in the section  “Cancer risks during Women’s Reproductive Years”

A review of principal research studies relating breast cancer risk to use of female hormones in the post menopausal period is shown in appended Table 3. Values here include risk ratios overall, error margins, and hormone risk per year of exposure. All but four of the risk ratios cited show an increased risk for breast cancer. Because the error margins of some of the studies are quite high, it is important to note the average values cited in the two meta analyses shown in italics. The very comprehensive meta study B17 cites an average value of 1.023 or a 2.3% increased risk of breast cancer for each year of hormone exposure from 51 different studies.  This value is accepted as best and is used in Life Ahead for actual years of post-menopause hormone use.  Note that the added risk of breast cancer per year of use dropped from 5.9% to 3.5% per year of exposure during the pre-menopausal years.  This post-menopausal value of 2.3% suggests a continuation of this decline of percentage risk at still older ages.

A probable widespread assumption by many doctors has been that survivors of breast cancer should not take post-menopausal hormones because they do increase risk for cancer-free women.  The actual data from two meta analyses cited following as B15 and B18 refute this speculation.  Surprisingly, each of these meta analyses found women survivors of breast cancer benefited from hormone use. Most of these studies were quite small individually and could have involved various problems. But the fact that nearly all of these studies found the same result individually, and each overall meta found a similar benefit should be reassuring to women wishing to use post-menopausal hormones.  A most surprising result was that of the meta D4 of 6 studies of hormone for use by cancer survivors in reducing death from all causes by 79%.

After hormone use is stopped, the risk of breast cancer moves back down over a period of several years to the risk of non-hormone users.  Life Ahead assumes that this risk reverts back to the non-user level 3 years after hormone use has been stopped.  No difference in the effect of Estrogen and Estrogen + Projesterone combinations was found for breast cancer. 

Endometrial (Uterine) Cancer:  Research on the effect of post-menopausal hormones on cancer of the uterus is included in Table 4. Hormones substantially increase the risk of uterine cancer. The risk ratios for use of Estrogen alone vary from 4 to 9 fold, with the highest value of 9 times for 10 or more years noted from  the study U1 meta analysis of 30 studies. Nearly all research studies showed that the effect of Estrogen on this cancer was related to years of use, and it seems apparent that estrogen develops risk as a function of estrogen times years. An average of the 7 best comparisons for estrogen therapy without projesterone was a risk value of 1.22 ^ years of use. This agrees closely with the specific estimate for this compound factor of 1.21 cited in study U6. This average result is well within the accuracy range of the clinical study U7 that included rather few cancer events. Life Ahead uses the value of 1.22 ^ years of exposure but with a maximum risk accepted of 8. Even recognizing that the usual risk of endometrial cancer is only about a quarter of that of breast cancer during post-menopausal years, this risk adds a serious negative for use of estrogen alone.

Progesterone has a major protective effect on endometrial cancer. The average of the results for use of usual estrogen-projesterone combinations is risk factor 1.06 ^ years-of-use.  This risk probably depends on the amount of Projesterone used.  The single result in study U4 found that continuous rather than 10 day per month use of projesterone accomplished a near complete protection against endometrial cancer.  Life Ahead assumes a risk ratio of 1.06 ^ years of use for estrogen with projesterone, with a maximum accepted risk of 1.40.

Life Ahead thus removes the risk of endometrial cancer if a hysterectomy is noted in the program entry screen. The risk of hormones declines after their use is stopped.  Life Ahead assumes risks revert to that of non-users three years after stopping estrogen-projesterone and five years after stopping estrogen alone therapy.

Ovarian Cancer:  Only two quantitatively useful studies of the effect of post-menopausal hormones on risk of ovarian cancer were located.  Both suggest that risk is related to years of use, and that this risk is modest at a level of about 1.05 ^ years of use. No difference in risk was noted in either of these studies for use of estrogen alone or estrogen + projesterone  Thus the value of 1.05 ^ years-of-use is used for estimating the increased risk of ovarian cancer as a result of post menopausal hormone use.  Only one useful study was found relating the risk of hormones to cervical cancer. This and other fragmentary data found suggest that the risk of hormones on this cancer is modest.

Genital Cancer:  The present Life Ahead version displays only risk levels for women’s genital cancer as a result of the sum of that at the above three cancer sites.  This is to minimize the complexity of program output. The usual proportions of these cancers at the important ages above 60 for women’s risk from hormones are endometrial, 58%; ovarian, 32%, and cervical, 10%.  Thus Life Ahead thus computes genital cancer risk as endometrial risk * 0.58 + ovarian risk * 0.32  +  cervical risk *0.1.   Because each of these values results from a times-years or power function, the life table risk is computed stepwise with increasing and differing risks for each year of exposure to hormones.  At identified times after hormone use is stopped the risk moves lower and eventually revert to the annual risks of no  hormone use.

Colorectal Cancer:  The key data used for colorectal cancer as a function of post-menopausal hormone use are summarized in Table 5 appended. The results for all studies are remarkably consistent.  Omitting the values of “Ever” use of hormone that include results for individuals that had stopped its use, the average risk ratio of 0.66 from the meta study C6 should provide an accurate average of results of the other 6 major individual studies cited.  No effect on this cancer for duration of use was found in either the specific studies that studied risks of its use, or in the overall data summarized in Table 5. Also the limited data available showed no difference between the benefits of estrogen or  estrogen + projesterone in reducing risk of colorectal cancer..

The beneficial effect of hormones on colorectal cancer is surprising.  Hormones promote risk of other cancers, and usually operate in the times-years mode expected of carcinogens.  Obviously a quite different mechanism is involved here that should be studied carefully.  The benefit of hormones on colorectal cancer helps offset part of the increase in risk the produced by other cancers. 

Death from All Causes:  Table 3 shows some values for Death from All Causes.  Studies D1 and D2 show substantial reductions in overall death rates for use of hormones for usual measured durations of its use. Study  D3 or the WHI study found little change in death risk for hormone use during the relatively short term of this study, and this result had a substantial margin of error. Follow-up reports on this study D3 as noted in the accompanying section on heart disease produced similar results.

Life Ahead Analysis of the use of Post-menopausal Female Hormones

Life Ahead implements all of the above risk valuations for each of the above diseases.  These risks usually differ not only for each disease but for each disease at each year of hormone use during life. The first tabulation following is from program computations for an average non-smoking US woman of age 50 that starts her hormone use at an age 49 time of menopause and continues this hormone use for life.  the symbol E identifies estrogen alone, and E + P denotes a usual estrogen-projesterone combination.  An assumption of no hormone use in this analysis produces for an assumed ‘average’ woman 10,556 Well Days (28.9 years) of healthful file ahead. Her use of E produces a small loss of 107 Well Days, but the use of  E + P produces a gain of 403 Well-Days of life.

But these results should be qualified further.  For the first 2-3 years of hormone use risks of both coronary disease and cancer risks are higher with use of hormones E or E + P.  But the excess risk to an individual during this time is quite small.  After this time of use, the risk of coronary disease gradually moves lower, and at about 5 years of E+P use effective overall health risk for use of these hormones becomes about break-even.  From 5 years use to 15 years use hormones E+P provide a small advantage in Well-Days.  During last years of life and hormone use for more than 15 years their use provides a larger advantage, and most of the above total benefit of 403 Well-Days.  There is a qualification on all of this because even with analyses of all available research on the risk of each disease, an error margin remains, and we do not have good data on hormone use by women much beyond age 65.  But on balance, E+P use for at least 5 years probably will not produce a significant excess in health risk. If hormones have been used for several years, it is unlikely that any advantage will be obtained by discontinuing their use.   For perspective, eating a half grapefruit each day for life will produce a larger benefit than will the 403 Well-Days computed for lifetime use of hormones. When compared with benefits potential from life habits as diet and exercise the risks or benefits involved with use of hormones are minimal.     

The table following shows that risks of cancer for an average  50 year woman to age 60 are much higher than are risks for heart disease.  The very high risk of the uterine component of genital cancer and somewhat increased risk for breast cancer outweighs benefits to coronary disease and results in a modest loss in Well Days for use of Estrogen alone. The long recognized role of progesterone in substantially reducing the risk of uterine cancer reverses this and produces an advantage of about 400 Well Days to healthful life.   But as perspective, the risks or benefits related to post-menopausal hormone use are small compared with even modest variations in lifestyle diet or exercise.

                      Woman of age 50                                                      ---- % Risks of Disease to 60 ---

Age Start  E or    Well    Diff in         Coronary   Breast  Genital All Causes 

Hormones   E+P     Days   Well-Days        H Disease  Cancer   Cancer   Cancer

  No Hormones     10,556    Base=0           3.42      4.13     1.40      9.27

  49       E      10,449     -107            1.94      4.81     4.36     12.43

  49     E + P    10,959     +403            1.94      4.81     2.01     10.07 

Many Life Ahead options were tried for women of varying ages and initial risks for both estrogen and estrogen-projesterone and for differing times of hormone use to explore how these varying combinations might affect life-time risk. Results usually confirmed those reached from the statistical analysis of past or pre-WHI research results on which most medical recommendations were then based. E+P usually develops better health than does E alone.  A family risk of cancer may remove computed health benefits for E + P.  But as noted previously, available research does not show that risk will be higher for women survivors of breast cancer than for those cancer-free.

Hormone use reduces risk of osteoporosis, and provide much comfort to some women.  The WHI study suggested that hormones produced a higher risk of dementia, but as for heart disease this could be a result of their short term and not a long term use.  On the negative side, use of hormones over life can involve an appreciable economic cost.  Life Ahead does not include use of hormones as a general health action that should be considered to improve lifetime Well-Days.  Use of hormones must be a medical decision - and not one of Wellness.  Thus benefits or debits from the use of hormones is not noted automatically in Life Ahead results. But a computed value of overall risk or benefit on each disease and for Well-Days will be included in the profile of any woman that uses post-menopausal hormones, and any woman can explore for herself the results computed by the program at her individual risks and habits for using various hormone options as type of hormone, and time of their use.

Women interested in obtaining a Life Ahead estimate of the risk of their use of hormones should download the free computer program from  Download Lifeahead, and enter present habits and factors in the eleven display screens accessed from the top option 'Enter or Edit Wellness Data'.  Although not required, an actual entry of a usual diet may improve slightly  the accuracy of the computation. Of particular importance to the computation are family risks of cancer and heart disease, and entries asked for in the female entry display. First run the program without entry of hormones to obtain a base result at user conditions.  Then ask from the result screen of this first computation to change 'Female Factors' and enter start age of hormone use and re-compute results.  The program will then display a change in risks of heart disease and cancer and change in Well-Days of life for hormone use over rest of life.  Alternates are to specify the an age at which hormones will be stopped, and type of hormone used. You can try and retry various entries to obtain estimates for difference use scenarios from the base or initial result obtained without use of hormones. Again, although Life Ahead may now produce a "State-of-Art' computation of risks of hormone use, the actual use of hormones can consider other factors and be a medical decision made by women with their doctor.

                                                     The TABLES of ACTUAL RESEARCH RESULTS

In using these tables, keep in mind that a risk ratio below 1.0 shows a beneficial effect of hormones.  A ratio of 0.5 denotes half the risk of a disease for hormone use as for non-use.  A ratio of 2.0 shows a twice higher risk associated with hormone use.

                                         Table 1  Female Hormones and Heart Disease

                       See table and analysis on   Female Hormones-Heart Disease Supplement

                                               Table 2:  Female Hormones and Stroke    

               Table 3:   Female Hormones and Death from All Causes

                        Table 4:   Female Hormones and Breast Cancer