Intended for healthcare professionals

Primary Care

Changes in use of hormone replacement therapy after the report from the Women's Health Initiative: cross sectional survey of users

BMJ 2003; 327 doi: (Published 09 October 2003) Cite this as: BMJ 2003;327:845
  1. Beverley Lawton, senior research fellow (bevlawton{at},
  2. Sally Rose, research fellow1,
  3. Deborah McLeod, research director1,
  4. Anthony Dowell, professor of general practice1
  1. 1 Department of General Practice, Wellington School of Medicine and Health Sciences, PO Box 7343, Wellington South, New Zealand
  1. Correspondence to: B Lawton
  • Accepted 18 September 2003


In 2002 the Women's Health Initiative reported the results of a study on the effects of combined oestrogen plus progestogen on healthy postmenopausal women. They found increased rates of breast cancer, coronary heart disease, stroke, and venous thromboembolism and decreased rates of hip fracture and colorectal disease but no “global” benefit. They concluded that combined oestrogen and progestogen is not suitable for the prevention of chronic diseases.1 The subsequent extensive media coverage came at a time when the prevalence,2 3 duration,2 4 and use of hormone replacement therapy (HRT) for the prevention of chronic disease had been increasing.2 4 After the report, government regulatory authorities issued advice to health professionals and women, and guidelines relating to the postmenopausal use of hormone replacement therapy were updated.5 We examined changes in HRT use since the publication of the report.

Participants, methods, and results

Between January 2000 and November 2002, 3500 women were screened from 23 general practices in four New Zealand centres as part of the recruitment process for the women's international study of long duration oestrogen after menopause (WISDOM) and a New Zealand observational study. We surveyed 998 women who were using HRT at the time they were screened but who were ineligible or unwilling to join the international study.

We sent participants an information sheet and questionnaire by post six months after the trial results were published. One reminder questionnaire was sent to non-responders, and 810 surveys (81%) were completed and returned. In total 776 respondents were taking HRT when the trial results were published and were eligible for inclusion. No significant differences were observed between non-responders (n = 188) and responders (n = 776) regarding age (χ2 = 2.89, df = 3, P = 0.41) and education (χ2 = 1.78, df = 3, P = 0.62). Non-responders were more likely than responders to have taken HRT for less than five years (χ2 = 9.71, df = 2, P = 0.008). The table shows the analyses for the 734 respondents for whom we had complete data on HRT use.

We analysed data using SAS Insight and estimated associations between a range of independent variables and the dependent variables (stopping and restarting HRT) from a generalised linear model with a log link and binomial errors. Independent variables included women's age at the time of the survey, type of HRT, years of HRT use, hysterectomy status, education, and reasons for starting HRT.

Of the 734 respondents, 423 (58%) stopped taking HRT. Of the 423 who stopped when the results were published, 132 (18%) had restarted at the time of our survey and 291 (40%) had not. Most respondents (610, 83%) reported that they had discussed HRT with a health professional. Older age, use of combined HRT, and longer duration of HRT were associated with stopping HRT. The association between stopping HRT and increased age was explained by the duration of HRT use (table).

Factors associated with stopping hormone replacement therapy in six months after oestrogen and progestogen trial report

View this table:

Restarting HRT was associated with taking oestrogen only, use for relief of symptoms, and having a hysterectomy. Of the 132 women who restarted, 100 did so because of the return of symptoms, 16 because they “felt better” on HRT, and 15 for other reasons.


After the publication of the results from the oestrogen plus progestogen trial there was a substantial change in HRT use among the women we surveyed: 58% initially stopped taking HRT and 18% of them subsequently restarted, leaving 40% stopped at the time of this survey. Data suggest that stopping HRT was consistent with updated international guidelines. Our survey has limitations as we did not use a random sample of HRT users, nor is the background rate of stopping HRT known.


We thank Selina Brown and the team of research nurses for their role in the collection of interview data from the cohort, for commenting on drafts of the questionnaire, and for assisting with sending the survey; Katie Abbott for entering questionnaire data on the database; Sonya White, who assisted with the questionnaire design; and the women for participating in this study.


  • Contributors All authors developed the idea for the study and participated in the questionnaire design. SR drafted the questionnaire, reviewed the literature, and coordinated the survey. BL and SR wrote the introduction, SR and DM carried out the analysis and interpretation of the data and wrote the participants, method, and results sections. All authors edited and were responsible for the final version of the paper. BL will act as the guarantor.

  • Funding The UK Medical Research Council (MRC) funded the recruitment of women into the WISDOM trial.

  • Competing interests BL is a member of a New Zealand Guidelines committee focusing on the appropriate prescribing of hormone replacement therapy and has separately received conference and research grants from pharmaceutical companies. BL and AD were the principal investigators and SR a coinvestigator of WISDOM in New Zealand.

  • Ethical approval The survey was approved by the Wellington Regional Ethics Committee, accredited by the Health Research Council of New Zealand.


View Abstract