Prof. John Studd. Women's Health Clinic
clinical gynaecologist
clinical gynaecologist

print this page


Update on Women's Health Initiative (WHI) study and Million Women's Study (MWS) on understanding of the safety of HRT

John Studd, DSc, MD, FRCOG
Consultant Gynaecologist, Chelsea & Westminster Hospital, London

Until two years ago HRT was quite straightforward. It stopped menopausal symptoms of hot flushes and vaginal dryness, improved depression and sexuality, had a protective influence on osteoporosis, heart attacks, strokes, colon cancer and probably Alzheimer's disease. There was probably a very slight increase in breast cancer but it was easy to believe that this was surveillance bias or the difficult precise pathological diagnosis in an organ stimulated with oestrogens (no independent review of breast pathology in these studies has ever taken place) as the mortality rate from breast cancer in these patients was much less than national averages in all papers. Women on HRT also lived an average 2.5 years longer than non-users.

These conclusions were the results of 30 years of clinical epidemiological and scientific data which seems to have been eclipsed by the conclusions of two large but greatly flawed studies.

The American WHI study of "healthy" 19,000 women confirmed the slight increase in breast cancer, the decrease in the hip fracture, the decrease in vertebral fracture and the decrease colon cancer but surprisingly revealed more heart attacks and more strokes. The mortality was not increased. Although this was meant to be a primary prevention study of the sort of menopausal women that we treat, the patients were of an average age 50-79, average age 63 with 23% of patients recruited over the age of 70. They were all given a standard dose of Premarin 0.625 mgs and MPA 2.5 mgs. The patients were overweight, 40% were hypertensive, 40% were receiving statins and 8% had had a previous heart attack. Thus the wrong patients of the wrong age group were given the wrong treatment and therefore the conclusions are very, very suspect. There were not enough patients in the 50-55 year group to make any observation about risks. This is regrettable that is the usual age when patients in this country commence HRT.

In fact looking at the raw data there is no increase in heart attacks year by year but in year 4 there was an unexplained drop of heart attacks and strokes in the placebo group which sprung the statistical significance thus stopping the oestrogen/progestogen arm of the trial.

A subsequent paper from WHI indicated that there was no improvement in quality of life with this HRT preparation. This was no surprise as the patients were asymptomatic as part of their inclusion criteria for the study. If there was no improvement in quality of life, there would be no problem but as women often have their lives transformed by the improvement of insomnia, depression, anxiety, dyspareunia etc., the personal dilemma for them is whether they should abandon this treatment. Regrettably this paper is now being referenced to indicate that HRT is not only dangerous but it doesn't even remove symptoms.

This has led to a huge decrease in the prescribing of HRT to the extent that one of the WHI investigators, Professor Susan Johnson, has publicly stated that the reaction to the WHI result has been too extreme.

To confuse things even more, it has been reported recently (04.03.04) that the oestrogen arm of the trial has now been stopped because the increase in strokes in this population has been confirmed. However, there is no increase in breast cancer or heart attacks in these 11,000 women studied over seven years.

It is no wonder that both doctors and patients are confused.

The Million Women Study was (if possible) even worse with more than 12 obvious errors and discrepancies in the text even to the table referring to ethinyloestradiol instead of oestradiol which was not picked up by the authors, clinical reviewers, or the editors. If these errors are above the surface, it brings into doubt the conclusions from the data collection and statistics under the surface. For example, the large increase in breast cancer after one year of oestrogen therapy is clearly undiagnosed breast cancer from the index mammogram as studies of the biology of breast cancer indicate that it takes five years before it is diagnosed as a 1.0 cm lump. Similarly the average time of diagnosis to death being 2.4 years is hard to believe as the average survival for metastatic breast disease is 3 years.

The many other faults, including the use of a single questionnaire only, and the exclusion of 4000 cancers from analysis can be found as a chapter on my website (

My view, and one shared by the British Menopause Society, would be that

  1. Oestrogens still have a place for menopausal symptoms with the lowest dose for the treatment of that particular symptom being used. The dose should, however, be high enough to solve the problem. For example, The dose for treatment of hot flushes would be less than for the treatment of hormone responsive depression
  2. Oestrogen therapy remains first line therapy for the prevention and treatment of osteoporosis.
  3. With our current state of knowledge oestrogens should not be used for the prevention of heart attacks, strokes or Alzheimer's disease. But there is no evidence that HRT as used in young menopausal women in the UK carries any extra risk of coronary heart disease or strokes.
  4. Consideration should be given to the use of non-oral routes as this avoids the entero-hepatic circulation and the production of excess coagulation factors from the liver.
  5. The need to give HRT should be reviewed annually and possibly long-term therapy with the reported extra risk of breast cancer (12 per 1000 at 15 years) should be avoided if possible.
  6. Until the breast cancer controversy is cleared up, it would be wise to recommend annual mammograms.

The problem is that it may take a month for bad news to be accepted but ten years to correct it particularly if a major Press conference occurs several days before publication. This occurred in the WHI study, the Million Women Study and, to put things in context, the MMR study. It does seem that the controversial the data the more likely that it is presented to the Press in this way so it is front page news before the scientific community has had time to analyse and interpret the publications. We have to find a more sensible way of communicating important scientific work to the public.


On the 2nd March, 2004 there was a Press Release from WHI to say that they had discontinued the oestrogen-only study because it confirmed the increase in strokes found in the earlier oestrogen/progestogen study. However, they found no increase in breast cancer and no increase in heart attacks! I agree that it is all very confusing.

The message from me is that we should avoid using Premarin with its adverse effect upon triglycerides, using oestradiol preferably by the transdermal route - patches, gels or implants - which does not stimulate coagulation factors from the liver.