Abstract
Lokkegaard and colleagues report a large national study of hormone therapy (HT) use and the risk of myocardial infarction (MI) [1]. Clinical use of HT in peri- or post-menopausal women has changed dramatically over the last 5 years because of increasing concerns regarding the risks of cardiovascular disease, cancers such as breast cancer, and cognitive decline or dementia. The cardiovascular risks were reviewed in this journal in 2007 [2]. Based on evidence from randomized, controlled trials (RCTs) reported by mid-2006, it was concluded that there was an overall increased risk of cardiovascular disease events in women using HT (four additional cases per 1000 HT users aged 50–59 years over 5 years, number needed to harm: 250; ten additional cases in women aged 60–69 years, number needed to harm: 100). The majority of excess cardiovascular events were venous thromboembolic events (VTE; deep vein thrombosis or pulmonary embolism); but there was also a significantly increased risk of stroke, especially severe stroke (fatal or disabling) [2]. Overall, there was no significant increased risk of coronary heart disease (CHD) including MI, although it was noted that there might be an excess of early CHD. It was also noted that there might be an increased risk of peripheral arterial disease (PAD) [2].
In 2008, two further meta-analyses have been published [3,4] that refine the estimates in this previous review [2]. Sare et al. reported an updated meta-analysis of 31 RCTs, which confirmed increased risks of VTE (Odds ratio [OR]:2.05; 95% confidence interval [CI]: 1.44–2.92), stroke (OR: 1.32; 95% CI: 1.14–1.53) and stroke severity (OR: 1.31; 95% CI: 1.12–1.54) [3]. By contrast, CHD events were not increased (OR: 1.02; 95% CI: 0.90–1.11). Although most trials included older subjects, age did not significantly affect risk. The addition of progestogens to estrogens doubled the risk of VTE [3]. In a further meta-analysis of VTE risk, Canonico et al. observed that the increased risk in HT users was significantly higher in users of oral preparations, compared with transdermal preparations [4].
In summary, while there is now convincing evidence from meta-analyses of RCTs that HT use significantly increases the overall risk of CVD, VTE and stroke, there is continuing uncertainty regarding the overall effects of HT on CHD and PAD, and also the effects of both user characteristics (e.g., age) and HT characteristics (e.g., unopposed estrogen vs estrogen–progestogen combinations, continuous vs cyclical preparations and oral vs non-oral administration). While the final results of RCTs, such as the Women's International Study of long duration estrogen after menopause (WISDOM) [5] and the Women's Health Initiative studies [6] are currently being published, there is a need to supplement their information with those of large observational studies.
Results
In an observational study published in a recent issue of the European Heart Journal, Lokkegaard et al. investigated the risk of MI as a result of HT, with a focus on the influence of age, duration of HT, various regimens and routes, progestogen type and estrogen dose. All healthy women in Denmark aged 51–69 years (n = 698,098) were followed between 1995 and 2001. A daily updated national capture of HT use was determined on the basis of a central prescription registry. National registers identified 4947 cases of incident MI.
Overall, this study observed no increased risk of MI, with a relative risk (RR) of 1.03 (95% CI: 0.95; 1.11) in women currently using HT compared with women who never used HT. The age-stratified RR appeared higher in women aged 51–54 years (RR = 1.24; 95% CI: 1.02–1.51) compared with women aged 55–59 (RR = 0.96; 95% CI: 1.02–1.51), 60–64 (95% CI: 0.97–1.27) or 65–69 years (RR = 0.92; 95% CI: 0.80–1.06). In women 60–69 years of age, the previous use of HT was associated with a decreased risk of MI.
Regarding the type of HT preparation, the highest risk of MI was found with continuous HT regimens (RR = 1.35; 95% CI: 1.18–1.53). No increased risk was found for unopposed estrogen, cyclic combined therapy or tibolone. A lower risk of MI was found for transdermal HT preparations compared with women who never used HT (RR = 0.62; 95% CI: 0.42–0.93) and compared with users of oral unopposed estrogen. No associations were found with progestogen type or estrogen dose. In women receiving combined therapy, no difference was detected between oral or transdermal administration. Vaginal estrogen was also associated with a lower risk of MI (RR = 0.56; 95% CI: 0.44–0.71).
Significance
While the associations of HT with cardiovascular disease in observational studies need to be considered with caution owing to potential biases, such as the healthy user bias [2], the strength of this study comes from its large size, and the national, unselected data from Denmark, a country that provides free access to medical care. Consequently, HT use is generally not associated with a healthy user lifestyle [1].
While this study found no overall increase in the risk of MI among HT users, an increased RR (of ~25%) was observed in younger women (aged 51–54 years). This finding renews concern regarding the suggestion that younger age may confer a low RR of CHD, as has been suggested by a nonsignificant trend reported in a recent analysis of the WHI data [6]. Therefore, at present, the overall evidence suggests that (even at a younger age) HT does not protect against CHD, may confer a small increased risk for CHD and, most importantly, significantly increases the risk of total cardiovascular events including VTE and stroke.
Given the increased risk of cardiovascular events, what HT preparations might be safest for women, especially those at a higher cardiovascular risk? The present study observed the highest risk in users of continuous combined HT – an effect that may be caused by increased tendency to thrombosis, a possibility supported by a recent meta-analysis of RCTs, which observed that combined HT doubled the risk of VTE compared with unopposed estrogen preparations [3]. No increased risk of MI was observed for other types of HT. The lower risk of MI observed with transdermal HT is interesting, because a recent meta-analysis observed a lower risk of VTE in users of transdermal HT [4], perhaps owing to the lesser systemic activation of blood coagulation [2,4]. This mechanism might also explain the observed lower risk of MI with vaginal HT.
Future perspective
To clarify the risks of different cardiovascular events in HT users by time, type of preparation and user characteristics, a collaborative meta-analysis of individual data in RCTs is required [2]. Such an analysis will be usefully supplemented by large observational studies such as that of Lokkegaard et al., as well as by postmarketing surveillance of newer HT preparations [1,2].
Executive summary
Hormone therapy significantly increases the overall risk of cardiovascular disease events (deep vein thrombosis, pulmonary embolism and stroke, especially severe stroke), but the risk of coronary heart disease, such as myocardial infarction (MI), is not well established.
Results of a Danish national study in women aged 51–69 years during 1995–2001 are consistent with the literature, in that there was no overall increase in risk of MI in current hormone replacement therapy users.
However, an increased risk of MI (by ~25%) was observed in younger women (aged 51–54 years). These findings support the recommendations in current evidence-based guidelines that hormone therapy (HT) use should generally be limited in duration.
The highest risk of MI was found with continuous HT regimens. No increased risk was found with unopposed estrogen, cyclical combined therapy or tibolone. A lower risk of MI was found with transdermal or transvaginal routes. No associations were found with progestogen type or estrogen dose. These findings should be considered (together with other evidence) in evidence-based guidelines, and in clinical practice, for selection of HT regimens, especially in women with increased cardiovascular risk.
Footnotes
Acknowledgements
Gordon Lowe thanks Helen Mosson for her assistance in preparing the manuscript.
The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.