Commentary: Hazards of studying women: the oestrogen oestrogen/progesterone dilemma

Lewis H Kuller

Department of Epidemiology, University of Pittsburgh, GSPH, 130 North Bellefield Avenue, Room 550, Pittsburgh, PA 15213, USA. E-mail: kullerl{at}edc.pitt.edu

The recent report from the Women's Health Initiative (WHI) of lack of benefit of oestrogen and progesterone (E+P) therapy in reducing the risk of coronary heart disease (CHD) has generated considerable discussion within the epidemiology community as to the importance of observational studies.1–6 The results of the WHI are not an isolated phenomenon. Rather, as has been noted, a series of clinical trials have documented lack of benefit of oestrogen alone or E+P therapy in both primary and secondary prevention of CHD as well as in the progression of carotid intimal medial thickness or coronary stenosis.7

Many women have been on long-term hormone therapy (HT), especially E+P, not necessarily because of menopausal symptomatology but as preventive therapy, especially for CHD. It has been estimated that approximately 7 million women in the US are on HT.1 The potential adverse effects of E+P, especially the increased risks of stroke, venous thromboembolism, breast cancer, and dementia as well as the lack of benefit in reducing the risk of CHD has had major public health implications.

Stampfer and Colditz, in their analysis of epidemiological evidence, made a substantial error in interpreting the existing data that contributed to the widespread use of an ineffective and perhaps hazardous therapy.8 Their review of the published data up to the time of 1991 is adequate and objective. They unfortunately dismissed the potential differences or biases associated with the use of oestrogen or E+P therapy that may have contributed to the benefits in observational studies. One of the weaknesses of their paper was their failure to determine why the women were on HT and especially to do a separate analysis for women who had an artificial menopause, i.e. oophorectomy, especially at earlier ages prior to menopause. The monkey studies, which strongly supported the benefits of E or E+P therapy in reducing the extent of atherosclerosis, were oophorectomy models combined with very high cholesterol or saturated fat diets9,10 substantially raising blood cholesterol. The WHI oestrogen only arm will include a fairly large number of women who have had an oophorectomy premenopausal and it is conceivable, but unlikely, that women with these characteristics will have some benefit in terms of reducing coronary artery disease from oestrogen therapy alone.

They noted that the observational studies did not prove causality but unfortunately then left the reader with the presumption that the results were so strong that they were likely to be causal, i.e. that oestrogen therapy for women of all ages, irrespective of prior oophorectomy and duration of therapy, was highly efficacious.

They focused part of their evaluation on the comparison of baseline characteristics of women on or not on HT. They did not consider the possibility that women on HT were more likely over time to adhere to other health behaviours that would reduce their risk of CHD, especially decrease in cigarette smoking, control of hypertension and lipid therapy, increase in physical activity, etc. over time. The marked differences in educational level by hormone use were also unfortunately not given much credence in their paper.11

It is important in evaluating the evidence to look at the totality of epidemiological studies and not focus only on the case-control and longitudinal studies within relatively homogenous populations. The fact that CHD incidence and death rates do not increase dramatically in populations which had very low utilization of HT perhaps would have been a clue that the importance of relative oestrogen deficiency among most postmenopausal women may not have been a major determinant of CHD. The increased risk of CHD in men on oestrogen therapy in the Coronary Drug Project,12 even though the dose of oestrogens was fairly high, was an important clue to the potential thrombogenic effects of HT.

There have been attempts to find excuses for the different results in clinical trials and observational studies. Epidemiology is, by definition, the study of epidemics in a population. The greatest strength of observational epidemiology is in determining risk among and within populations that vary substantially in incidence and mortality from disease, either over time, place, or characteristics of individuals within a community. Observational epidemiological studies within relatively homogenous populations depend on the ability to carefully measure the independent and dependent variables accurately, repeatably, and in an unbiased fashion; for example, measurement of serum cholesterol and blood pressure levels and risk of CHD and stroke, relationship of cigarette smoking to risk of lung cancer, etc.

The exposure to hormones can be measured accurately. The rationale why some women do and others do not receive HT was not adequately determined nor other host characteristics, as well as the changes in other key risk factors after starting therapy that are independent of the specific effect of the therapy.

It is very possible that in the past, doctors selected women for HT because they were relatively oestrogen deficient, i.e. oophorectomy, symptomatology at time of menopause. Such women might benefit from HT as compared with similar women who did not receive HT. Once HT was provided to postmenopausal women irrespective of menopausal symptomatology, the adverse effects likely reduced or eliminated any benefits of HT. Epidemiology is the study of host, agent, and environment. Buyer beware for the epidemiologist that does not measure these variables and the interactions as carefully as possible.


    References
 Top
 References
 
1 Writing Group for the Women's Health Initiative Investigators. Risks and benefits of estrogen plus progestin in healthy postmenopausal women. Principal results from the Women's Health Initiative randomized controlled trial. JAMA 2002;288:321–33.[Abstract/Free Full Text]

2 Manson JE, Hsia J, Johnson KC et al. for the Women's Health Initiative Investigators. Estrogen plus progestin and the risk of coronary heart disease. N Engl J Med 2003;349:523–34.[Abstract/Free Full Text]

3 Harrington DM, Howard TD. From presumed benefits to potential harm—hormone therapy and heart disease. N Engl J Med 2003;349:519–21.[Free Full Text]

4 Whittmore AS, McGuire V. Observational studies and randomized studies of hormone replacement therapy: what can we learn from them? Epidemiology 2003;14:8–10.[CrossRef][ISI][Medline]

5 Grodstein F, Clarkson TB, Manson JE. Understanding the divergent data on postmenopausal hormone therapy. N Engl J Med 2003;348:645–50.[Free Full Text]

6 US Preventive Services Task Force. Postmenopausal hormone replacement therapy for the primary prevention of chronic conditions: recommendations and rationale. Ann Intern Med 2002;137:834–39.[Abstract/Free Full Text]

7 Kuller LH. Hormone replacement therapy and risk of cardiovascular disease. Implications of the Women's Health Initiative. Arterioscler Thromb Vasc Biol 2002;23:11–16.[ISI]

8 Stampfer MJ, Colditz GA. Estrogen replacement therapy and coronary heart disease: a quantitative assessment of the epidemiologic evidence. Prev Med 1991;20:47–63. (Reprinted Int J Epidemiol 2004;33:445–53.)[ISI][Medline]

9 Clarkson TB. Progression and regression of nonhuman primate coronary artery atherosclerosis: considerations of experimental design. In: Malinow MR, Blaton VH (eds). Regression of Atherosclerotic Lesions. New York: Plenum, 1984, pp. 43–60.

10 Clarkson TB, Anthony MS, Wagner JD. A comparison of tibolone and conjugated equine estrogens effects on coronary artery atherosclerosis and bone density of postmenopausal monkeys. J Clin Endocrinol Metab 2001;86:5396–404.[Abstract/Free Full Text]

11 Matthews KA, Kuller LH, Wing RR, Meilahn EN, Plantinga P. Prior to use of estrogen replacement therapy, are users healthier than nonusers? Am J Epidemiol 1996;143:971–78.[Abstract]

12 The Coronary Drug Project Research Group. The Coronary Drug Project: findings leading to discontinuation of the 2.5 mg/day estrogen group. JAMA 1973;226:652–57.[CrossRef][ISI][Medline]