The effect of menopausal transition on women's lives is complex and includes changes in physical health, psychosomatic domains, and personal life 1. Health-related quality of life may be severely compromised in women with vasomotor symptoms 2. Up to 40% of women in Sweden experience vasomotor symptoms until the age of 64 years 3. Postmenopausal hormone therapy has for a long time been the recommended first choice to alleviate hot flushes and night sweats 4567. Hormone therapy may have side effects such as vaginal bleeding, breast tenderness, migraine headaches, mood alterations, and abdominal bloating that may affect quality of life 8. Besides vasomotor symptoms, sleep and sexual problems are related to postmenopausal status 910.

Population-based studies indicate that hormone therapy improves vasomotor symptoms and sexual problems 11, but not quality of life 12. In the Heart and Estrogen/Progestin Replacement Study (HERS), women with hot flushes had improvements in emotional measures of quality of life 13. In the Women's Health Initiative (WHI) Trial, combined hormone therapy relieved some menopausal symptoms, but contributed to side effects 14, and no clinically meaningful effect on health-related quality of life was found 1516.

The Estonian Postmenopausal Hormone Therapy (EPHT) Trial was a long-term preventive trial of hormone therapy among healthy postmenopausal women. The EPHT Trial consisted of two sub-trials: one blind and one non-blind. Women were randomly allocated to combined continuous hormone therapy or non-treatment, or to blind hormone therapy or placebo. Their baseline health indicators showed better health than those of the participants in the WHI or HERS 17. At baseline, women in the EPHT Trial were aged on average 58 and were 8 years postmenopausal compared to an average age of 63 years (13 years postmenopausal) in the WHI Trial 14 and 67 years (18 years postmenopausal) in the HERS Trial 1318.

Besides studying the effect of postmenopausal hormone therapy on the risk of cancer, cardiovascular diseases and bone fractures 17 and the related use of health services 19, the other major aims of the EPHT Trial were to analyze the effect of hormone therapy on symptom reporting and health-related quality of life.

Symptoms reported most often by all participants at recruitment were aches or stiffness in joints (56%), chronic fatigue (53%), sweating (47%), hot flushes (45%), and backache (40%). There were no baseline differences in the prevalence of symptoms in different arms except for sweating, which was reported more often by women in the hormone therapy arm (50%) than in the non-treatment arm (44%) of the non-blind sub-trial (Table 2).

Longitudinal data analysis showed that throughout the trial, less women reported hot flushes (OR 0.20; 95% CI: 0.14–0.28), sweating (OR 0.56; 95% CI: 0.44–0.72) and sleeping problems (OR 0.66; 95% CI: 0.52–0.84) in hormone therapy arms than in non-therapy arms. Prevalence of hot flushes in different trial arms over time is presented in Figure 2, prevalence of sweating in Figure 3, and prevalence of sleeping problems in Figure 4. The difference between the trial arms in reporting sleeping problems at the end of the trial was less evident than at the end of the second trial year probably due to decrease in adherence rates. There was no difference between the treatment and non-treatment arms in reporting depression or any other symptoms (Table 2).

Longitudinal data analysis showed that hormone therapy increased the risk of vaginal bleeding substantially (OR 19.65; 95% CI: 12.15–31.79). More women in hormone therapy arms than in non-therapy arms reported vaginal bleeding in the previous 12 months both at the end of the second trial year (58 versus 13) and at the end of the trial (41 versus 13). The mean number of bleeding episodes per woman reporting bleeding in the previous 12 months was also higher in hormone therapy arms both at the end of the second trial year (3.7 versus 2.1) and at the end of the trial (3.5 versus 1.2). Bleeding episodes were considered to have been heavy by more women in hormone therapy arms than in non-therapy arms at the end of the second trial year (3 versus 0) and at the end of the trial (4 versus 0).

At the end of the first trial year, less women in hormone therapy arms (5%) than in non-therapy arms (9%) reported painful intercourse (OR 0.59; 95% CI:0.37–0.94), but more women complained of painful intercourse in hormone therapy arms (6%) than in non-therapy arms (2%) at the end of the trial (OR 3.12; 95% CI:1.65–6.25). The proportion of women having had no intercourse was not significantly different in different trial arms over time.

According to the EQ-5D questions, there was no difference between the arms in reporting problems with self-care, usual activities, pain or discomfort, anxiety or depression, but problems with mobility were reported more often by women in the blind hormone therapy arm (35% of respondents) than by women in the placebo arm (27% of respondents) at the end of the trial (OR 1.43; 95% CI 1.01–2.05). Despite the differences in reporting symptoms and problems with mobility, there was no difference in the distribution of women with different EQ-5D scores in trial arms at the end of the second trial year and at the end of the trial. As the distribution of women with different quality of life scores was strongly skewed, the results are presented in quartiles (Table 3). For half of the women in all trial arms, EQ-5D score was 0.90 at the end of the second trial year and 0.80 at the end of the trial.

Longitudinal data analysis showed that quality of life decreased with age (estimated difference -0.027 per one decade, SE = 0.009) and time (estimated difference -0.017 per one trial year, SE = 0.0024), but was not influenced by treatment or blinding.

Our data showed a decrease in reporting vasomotor symptoms and sleeping problems in hormone therapy arms, and an increase in the rates of vaginal bleeding over the observed period of time. Despite the difference in symptom reporting, no difference in quality of life scores between different trial arms was observed.

Hormone therapy is not used only around menopause. Before the publication of the results from the Women's Health Initiative Trial and the Million Women Study, it was estimated to be used by a third of postmenopausal women in Sweden 24, by forty percent in the United States 25 and by more than forty percent in Finland 26 for up to fifteen years after menopause. After publication of the findings from these studies the use of hormone therapy declined 2728293031, but still hormone therapy is used by tens of millions of women all over the world.

Women interested in participating in the trial were younger and more frequently reported depression and trouble sleeping than women not interested in participation. There was no difference in other socioeconomic characteristics and the reporting of other symptoms inquired about in the recruitment questionnaire 32.

We assume the higher proportion of women reporting sweats in the non-blind hormone therapy arm at recruitment to be a chance finding as treatment allocation was not known before recruitment. No significant difference in the numbers of women reporting sweats at baseline was observed between hormone therapy and non-treatment arms after combining data from both sub-trials. Therefore, the difference was not taken into account in the statistical analysis.

The proportion of women reporting problems in different trial arms at recruitment was higher than during the trial, probably because not all women with symptoms at recruitment responded to annual questionnaires. As the response rates to the annual questionnaires varied and the final questionnaire was completed at a different time, longitudinal data analysis was used to estimate the effect of hormone therapy on the prevalence of symptoms and quality of life over time.

No data for EQ-5D was collected at baseline. We assume that the EQ-5D scores at baseline were similar in all trial arms. Even though EQ-5D asks about mobility, self-care, usual activities, pain and anxiety or depression and is said to be one of the bluntest health-status instruments available 33, EQ-5D has been found to be suitable for measuring health-related quality of life among gynecological patients in all age groups 34. It is more likely to elicit a response than complicated measurement tools 35.

The beneficial effect of hormone therapy on vasomotor symptoms is well established 4567 as well as bleeding as its side-effect 11. The effect of hormone therapy on sleep disturbance has earlier been reported over a shorter period of time 1536. Although transition to menopause has been found to be associated with depressed mood 37, we found no difference in the proportion of women reporting depression in different trial arms.

Low adherence rates in the treatment arms may have diluted the effect of hormone therapy. The inconsistent effect of hormone therapy on painful intercourse during the EPHT Trial could be explained by lower adherence rates at the end of the trial in comparison with the adherence rates in hormone therapy arms at the end of the first trial year. Women were not persuaded to continue trial treatment, and the fall in adherence rates was the same as in real life circumstances 38.

Overall quality of life scores decreased with increasing age. The problems with mobility reported more often by women in hormone therapy arms in the EQ-5D questionnaire at the end of the trial may be a chance finding. The results are presented separately for both the blind and the non-blind sub-trial. Blinding had no effect on symptom reporting or quality of life over time.

In the WHI Trial, moderate to severe hot flushes were reported by 24% of participants aged 50–54 years and 15% for women aged 55–59 years at baseline, and night sweats by 21% and 14% accordingly 14. In the HERS Trial, hot flushes were reported by 16% of participants at baseline 13. In the EPHT Trial, the proportion of women with vasomotor symptoms was higher (Table 2).

In the WHI Trial, changes in menopausal symptoms were analyzed at the end of the first trial year 14, and quality of life was measured at baseline, at year one, and in a sub-group of women in year three 15. In the HERS Trial, quality of life and depressive symptoms were measured at a three-year follow-up 13. In the EPHT Trial, data on symptom reporting was collected annually, data on quality of life at year two and at the end of the trial, the mean follow-up period being 3.6 years.

In the WHI Trial, quality of life was assessed with the use of the RAND-36 Health Survey, and in addition data about depressive symptoms, sleep disturbance, sexual and cognitive functioning, and menopausal symptoms were collected 15. In the HERS Trial, physical activity was measured by the Duke Activity Status Index, energy and mental health by RAND scales, and depressive symptoms on the Burnam scale 13.

According to the WHI Trial data, combined hormone therapy improved vasomotor symptoms, vaginal or genital dryness, and all pain symptoms, but increased the rates of breast tenderness during the first year of use and vaginal bleeding persistently 14. No clinically meaningful effect on health-related quality of life was detected 15. The HERS trial data suggested that quality of life generally declined during the follow-up and hormone therapy had negative effects on physical function, but it improved depressive symptoms and quality of life for women with menopausal symptoms 13.

The follow-up period for symptom reporting and quality of life in the EPHT Trial was longer than in the WHI and HERS trials, while the regimen used was the same. The difference in outcomes can be explained by the difference in the participants' health indicators and age, by the difference in measurement tools, and by cross-cultural differences 39.

Data from the EPHT Trial show that postmenopausal hormone therapy decreased hot flushes, sweating, and sleep disorders, but increased episodes of vaginal bleeding. No effect on quality of life was observed. Physicians prescribing hormone therapy for postmenopausal women should take into account the complex of the effects of hormone therapy.

Health-related quality of life may also be influenced by cultural and socio-economic factors 40, therefore all tools that measure quality of life are highly influenced by personality and social circumstances 41. Which factors influence the quality of life in the ageing population in different cultural contexts deserves additional research, and measurement tools for quality of life could be further developed. The underlying mechanisms for subjective symptoms among postmenopausal women require future clinical research.

The trial was supported by the Academy of Finland (grants 69838 and 201490), STAKES (National Research and Development Centre for Welfare and Health), Finland, the Estonian Science Foundation (grants 5203 and 6570) and the Estonian Ministry of Education and Research (target funding 0192112s02). Drugs were donated by Wyeth Ayerst Company via the Women's International Study of Long Duration Oestrogen After Menopause (WISDOM) in the United Kingdom.

The funding bodies had no role in the study design, data collection, data analysis, data interpretation, writing of the report or in the decision to submit the paper for publication.

MR and EH contributed to the design of the trial, the development of questionnaires and to the manuscript; HK and SLH contributed to the development of questionnaires and to the manuscript; KF contributed to the data analysis and to the manuscript; PV contributed to the development of questionnaires, data collection and analysis, and to writing the manuscript.