Postmenopausal Hormone Therapy and Primary Prevention of Cardiovascular Disease—Nurses’ Health Study 20-Year Follow-Up
abstract & commentary
The 20-year follow-up from the nurses’ health study reports a reduced risk of coronary heart disease in current users of postmenopausal hormone therapy, but a slight increase in the risk for nonfatal stroke.
Grodstein and colleagues from the Nurses’ Health Study report the effect of postmenopausal hormone therapy on cardiovascular disease, based upon follow-up from 70,533 postmenopausal women. This prospective, cohort study recorded 953 myocardial infarctions, 305 coronary deaths, and 767 strokes from 1976-1996. The major observations are presented in the following tables:
| Table 1-Coronary Heart Disease | ||
| Cases | Adjusted RR (Confidence Interval) | |
| Current users | 259 | 0.61 (0.52-0.71) |
| 0.3 mg | 19 | 0.58 (0.37-0.92) |
| 0.625 mg | 9 | 0.54 (0.44-0.67) |
| 1.25 mg | 41 | 0.70 (0.51-0.97) |
| Estrogen alone | ? | 0.55 (0.45-0.68) |
| E + P* | ? | 0.64 (0.49-0.85) |
| Table 2-Stroke | ||
| Cases | Adjusted RR (Confidence Interval) | |
| Current users | 238 | 1.13 (0.94-1.35) |
| 0.3 mg | 9 | 0.54 (0.28-1.06) |
| 0.625 mg | 124 | 1.35 (1.08-1.68) |
| 1.25 mg | 46 | 1.63 (1.18-2.26) |
| Estrogen alone | ? | 1.18 (0.95-1.46) |
| E + P* | ? | 1.45 (1.10-1.92) |
| * Estrogen and Progestin together | ||
| Table 3-Fatal Stroke | ||
| Adjusted RR (Confidence Interval) | ||
| 0.625 mg | 1.01 (0.59-1.71) | |
| 1.25 mg | 1.25 (0.57-2.77) | |
| Estrogen alone | 0.81 (0.49-1.34) | |
| E + P | 1.22 (0.65-2.28) | |
These numbers indicate that current users of post- menopausal hormone therapy have a significant reduction in the risk of coronary heart disease with estrogen alone or with a combination of estrogen and progestin (although not reported, this is most likely nearly all sequential regimens). The results were adjusted for body mass index, diabetes, hypertension, elevated cholesterol, smoking, and age of menopause. Further analysis of diet, physical activity, and use of aspirin and vitamin supplements did not significantly change the results. Overall, there was no significant effect of hormone therapy on the risk of nonfatal and a fatal stroke. However, there was a suggestion that higher doses of estrogen and estrogen combined with progestin modestly increased the risk of nonfatal stroke (a statistically significant effect was present only with ischemic stroke, not with hemorrhagic stroke), but no significant increase in the risk of fatal stroke. (Grodstein F, et al. Ann Intern Med 2000; 133:933-941)
COMMENT BY LEON SPEROFF, MD
This latest update from the Nurses’ Health Study on the effect of postmenopausal hormone therapy on the primary prevention of cardiovascular disease provides no major changes from the 16-year report, about a 40% reduced risk for coronary heart disease, but it allows a better assessment of the effect of dose and duration of use.1 Overall, this report from the Nurses’ Health Study provides support for the belief that post- menopausal hormone therapy provides primary pre- vention against coronary heart disease, and that the doses of 0.3 mg and 0.625 mg of conjugated estrogens produce comparable effects.
Unfortunately, the report does not provide the case numbers for the relative risks associated with estrogen alone compared with a combination of estrogen and progestin, and for the analysis of fatal stroke. The tight confidence intervals suggest that the estimates of risk are precise, usually a result of adequate case numbers. However, notice that the stroke risk associated with 0.3 mg conjugated estrogens is based on only nine cases. Thus, one appropriate concern is whether the conclusions regarding combined estrogen and progestin are limited by small case numbers. It would also be of great interest to know whether the stroke results in the Nurses’ Health Study are influenced by the ages of the women. An obvious concern is whether the dose of estrogen should be decreased with increasing age.
In contrast to the uniform results from observational studies of the association between postmenopausal hormone therapy and coronary heart disease, epidemiologic data over the last 20 years regarding estrogen use and stroke have not been consistent. Many studies have indicated either no effect of postmenopausal hormone therapy on the risk of stroke or a reduction in risk associ-ated with estrogen or estrogen-progestin use.1-12
In a large Danish case-control study, no affect could be detected of either estrogen or combined estrogen and progestin on the risk of nonfatal stroke, both thromboembolic and hemorrhagic.11 A case-control study from Seattle found about a 50% reduced risk of subarachnoid hemorrhage with the use of postmenopausal hormone therapy, and the effect was even greater among smokers.10 In the prospective study of the Leisure World cohort, estrogen therapy was associated with a 46% overall reduction in the risk of death from stroke, with a 79% reduction in recent users.5 The population-based cohort study in Uppsala, Sweden, documented a 30% reduced incidence of stroke in postmenopausal users of estrogen, and, importantly, women prescribed an estrogen-progestin combination, containing a significant dose of the potent androgenic agent levonorgestrel, also experienced a reduced incidence of stroke.9 A reduced risk for mortality from stroke in this Swedish study was confined to intracerebral hemorrhage.13
Within this confusing mixture of results, there has been one consistent observation. The cohort studies (with a sufficient number of cases) that have assessed the effect of hormone use on the risk of death from stroke have all indicated a beneficial effect (except for the Nurses’ Health Study). For example, the National Health and Nutrition Exami- nation Survey (NHANES) recruited a large cohort of women from 1971-1975 for epidemiologic analysis. The follow-up longitudinal study of this cohort yielded a U.S. national sample of 1910 white, postmenopausal women. Postmenopausal hormone use in this cohort provided a 31% reduction in stroke incidence and a strongly significant 63% reduction in stroke mortality.8
One emphasis in the discussion in the current report from the Nurses’ Health Study was especially disturbing to me. Grodstein et al twice refer to their examination of women with previous coronary disease in the Nurses’ Health Study, concluding that short-term hormone use increased the rate of recurrent cardiac events, supporting the HERS Trial. Grodstein et al conveniently do not provide their numbers, but they have been presented in abstract form14 (See Table 4).
| Table 4 | ||
| Cases | Risk of Recurrent Cardiac Event | |
| Current Use | ? | 0.65 (0.45-0.95) |
| < 1 year | 6 | 2.1 (0.88-4.97) |
| 1-1.9 yrs | 3 | 1.01 (0.31-3.27) |
| 2+ yrs | 33 | 0.56 (0.37-0.85) |
These numbers do not indicate a statistically significant increase with less than two years of hormone use in women with coronary heart disease, and citing the results with short-term use, based on only six cases, is yet another example of selective reporting by the Nurses’ Health Study authors. Indeed, where there is some strength of numbers, these data support the idea that with increasing duration of exposure there is secondary prevention of recurrent events (as also noted in the HERS Trial)!
The editorial accompanying the report from the Nurses’ Health Study concludes that the disappointing results from the HERS Trial, the ERA Trial, and the recent report from the Women’s Health Initiative (WHI) indicate that clinicians should not use hormone therapy for the prevention of coronary disease until we have final data from randomized trials.15 I disagree with this conclusion. This latest report from the Nurses’ Health Study continues to provide support for the belief that postmenopausal hormone therapy provides protection against coronary heart disease in current users. It also suggests that we should begin to consider the use of lower doses of estrogen in older women (after age 65). The HERS Trial and the WHI report indicated a growing benefit with increasing duration of use, not a totally null or adverse effect. Randomized trial data will not emerge until 2006-2008. Until then, the current state of knowledge is sufficient, in my view, to support hor- mone therapy for primary prevention of coronary heart disease.
References
1. Grodstein F, et al. N Engl J Med 1996;335:453-461.
2. Petitti DB, et al. JAMA 1979;242:1150-1154.
3. Hammond CB, et al. Am J Obstet Gynecol 1979;133:525-536.
4. Bush TL, Barrett-Conner E, Cowan DK. Circulation 1987;75:1102-1109.
5. Paganini-Hill A, Ross RK, Henderson BE. BMJ 1988;297:519-522.
6. Thompson SG, Meade TW, Greenberg G. J Epidemiol Comm Health 1989;43:173-178.
7. Hunt K, Vessey M, McPherson K. Br J Obstet Gynaecol 1990;97:1080.
8. Finucane FF, et al. Arch Intern Med 1993;153:73-79.
9. Falkenborn M, et al. Arch Intern Med 1993;153: 1201-1209.
10. Longstreth WT, et al. Ann Intern Med 1994;121:168.
11. Pedersen AT, et al. Lancet 1997;350:1277-1283.
12. Petitti DB, et al. Stroke 1998;29:23-28.
13. Schairer C, et al. Epidemiology 1997;8:59-65.
14. Grodstein F, Manson JE, Stampfer MJ. Circulation 2000;100(Suppl):18.
15. Grady D, Hulley SB. Ann Intern Med 2000;133: 999-1001.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content