Editor's Note: Commentary based on Gudmundsson A, Aspelund T, Sigurdsson G, et al. Long-term hormone replacement therapy is associated with low coronary artery calcium levels in a cohort of older women: the Age, Gene/Environment Susceptibility-Reykjavik Study. J Am Geriatr Soc 2017;65:200-6.

Rationale for Study/Background: Although onset of cardiovascular disease in women occurs 10 years later than in men, it is still the number one killer. The role of hormone replacement therapy (HRT) in cardiovascular (CV) risk reduction has long been debated. Despite the negative findings in the Women's Health Initiative (WHI) overall, and resulting abrupt cessation of clinical HRT prescriptions thereafter, it was also the case that the WHI subgroup who started HRT within 10 years of menopause seemed to have better CV outcomes. Also, the 2012 study of young women (mean age 51) randomized HRT for osteoporosis prevention in Denmark also demonstrated better CV outcomes with HRT. Coronary calcium score is an established marker of atherosclerosis. The Reykjavik Age, Gene/Environment Susceptibility (AGES) population had longitudinal information on lifestyle, HRT use, risk factors, and calcium scoring with which to explore the association between HRT and coronary artery calcification (CAC).

Funding: Supported by National Institutes of Health (NIH/NIA), The Icelandic Heart Association, and the Althing (Icelandic parliament).

Methods

Design: Large population-based cohort study of older Icelandic women (ages 67-93) in the AGES-Reykjavik Study between 2002 and 2006, with follow up in the Icelandic Heart Association Registry.

Inclusion criteria: Participants with CAC testing on CT scan and mid-life physical exam at age <65 years with completed lifestyle questionnaire (age at menopause,  use of HRT, past and present physical activity, and number of hours of exercise reported per week).

Exclusion criteria: Participants missing registry or CAC requirements for the cohort.

Exposure: The timing of initiation and duration of HRT with propensity adjusted use of HRT in late life based upon covariates from midlife data.

Primary outcome(s): Median values of late-life CAC according to age group and HRT use.

Secondary outcomes: Prior history of coronary events based on a national registry for myocardial infarction and procedures including percutaneous angioplasty and coronary bypass surgery.

Statistical analysis: Multiple regression estimates of median CAC according to duration of HRT use (never, 0-5, 6-15, ≥16 years) and HRT initiation in relation to menopause (never, before menopause, within 1 year after last menses [menopause year], 1-5 years after menopause, >5 years after menopause).

Results: Of the 2,867 participants, 872 (30.4%) had used HRT, including 312 (10.9%) who were current users. HRT users were an average 2.6 years younger, had higher educational level, more physical activity reported, lower LDL-C, and lower BMI than never-users. The history of prior CV events was similar. CAC increased with age, with reduction among prior HRT users, and consistently lower CAC scores for current HRT users of all ages. There was an association between lower CAC scores and longer duration of HRT use, with a difference in median CAC of approximately 50% between the groups of never HRT and 16 or more years of HRT (P < .001). Women who started HRT in the menopausal year had the lowest CAC scores, significantly lower than women who never used HRT by 87 Agatston units (95% CI = 42-132, P ≤ .001). Women who initiated HRT later also had lower CAC scores than never users, by 56 Agatston units (95% CI = 13–98, P = .009).

Conclusion: This study found a strong association between duration of HRT use and lower CAC levels with the lowest CAC levels observed in those starting HRT at the time of menopause.

Limitations of study: This study was observational and residual confounding is likely. The outcomes of CAC score are a surrogate for CV events, and no differences in actual CV events in any group were noted. The cross-sectional nature of the data may result in loss of data due to drop out. Finally, HRT cessation may be due to other factors associated with CV risk.

Geriatric perspective for the cardiovascular clinician: This observational study adds to the supporting evidence of a positive effect of HRT on lessening atherosclerosis as measured by CAC. Yet it is far from actionable in clinical practice when added to the totality of the higher quality evidence on clinical outcomes. A similar association between HRT use and atherosclerosis (CAC) was noted among women 50 to 59 years old in WHI assigned to estrogen who had less calcified-plaque burden in the coronary arteries after trial completion than those assigned to placebo. The finding of better CAC scores when HRT is started at the time of menopause is also consistent with data suggesting less increased CV risk in younger women from HERS and WHI. The association between longer use of HRT and less CAC, a novel finding in this study, also remains confounded by a healthy user bias. Estrogen has complex biologic effects which may influence cardiovascular risk through multiple pathways. In this study, despite lower CAC burden with HRT, there was no corresponding evidence of a lower CV event rate in any HRT use group. Therefore, the prevailing caution against HRT use in post-menopausal women for CV prevention stands, even if support for a mechanistic hypothesis between estrogen and CAC is strengthened.

Keywords: Angioplasty, Atherosclerosis, Biological Products, Body Mass Index, Cardiovascular Diseases, Coronary Artery Bypass, Coronary Artery Disease, Cross-Sectional Studies, Estrogens, Exercise, Geriatrics, Hormone Replacement Therapy, Life Style, Menopause, Myocardial Infarction, Osteoporosis, Registries, Risk Factors, Risk Reduction Behavior, Tomography, X-Ray Computed

< Back to Listings