Prevalence and characteristics associated with use of hormone replacement therapy in Britain

Objective To describe trends in the prevalence of hormone replacement therapy (HRT) use among women living in Britain between 1981 and 1990, and to determine whether women who receive HRT are different from those who do not. Design Population-based, prospective, observational study. Setting Fourteen hundred general practices throughout the United Kingdom contributing data to the Oral Contraception Study of the Royal College of General Practitioners (RCGP). Population Prevalence of use: women still under general practitioner observation in the Oral Contraception Study at the end of December 1981 (n = 19,949), 1984 (n = 18,037), 1987 (n = 16,063), 1990 (n = 13,379). Characteristics of users: 3806 HRT users and 3806 never-users matched for age, hysterectomy status and duration of observation in the Oral Contraception Study. Main outcome measure Ever-use of hormone replacement therapy. Between 1981 and 1990 there was a threefold increase in HRT use among women both among those who had and those who had not had a hysterectomy. Increasing trends were apparent in all age groups. By December 1990, 19% of all women (36% of those who had had a hysterectomy and 16% of those who had not had a hysterectomy) had ever used HRT; 9% were classified as current users. Forty-one percent of women using HRT who had had a hysterectomy received combined (oestrogen and progestogen) preparations; 31% of those who had not had a hysterectomy received at least one prescription for unopposed oestrogen. Among those who had had a hysterectomy, HRT was more likely to be prescribed in women with a history of smoking, nonpsychotic psychiatric illness, hot flushes, other menopausal symptoms, oophorectomy, migraine or headache. Women with breast cancer were less likely to receive HRT. In women who had not had a hysterectomy, smoking, nonpsychotic psychiatric illness, hot flushes, other menopausal symptoms, migraine, headache and previous use of oral contraceptives increased the chances of HRT being used; a history of breast cancer, ischaemic heart disease or diabetes mellitus reduced the chances. Weaker associations of reduced risk were also found for hypertension and lower social class. The influence of hot flushes and other menopausal symptoms appeared to be stronger in women who used HRT before the age of 50, than in older women. Conclusions HRT use increased dramatically within the cohort during the nine year period. HRT users are different from nonusers in a number of ways which could have opposing effects on a woman's risk of cardiovascular disease.