IJE Advance Access originally published online on March 31, 2006
International Journal of Epidemiology 2006 35(3):738-739; doi:10.1093/ije/dyl039
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Commentary |
Commentary: Hormones, heart disease, and the definition of hormone ‘initiation’
Department of Epidemiology, Harvard School of Public Health, Boston, MA 02115, USA. E-mail: stampfer{at}hsph.harvard.edu
Observational studies consistently find that users of post-menopausal hormone therapy (HT) are at substantially lower risk of coronary heart disease (CHD).1 These results were replicated most recently by Kim et al.,2 in this issue, who find a 24% reduction in risk of myocardial infarction among current HT users. In stark contrast, randomized trials of HT show no overall reduction in CHD and provide convincing evidence for a short-term increase in risk when initiated among women with CHD and possibly among women with no clinical CHD.3 The apparent discrepancy has drawn appropriate scrutiny, and a variety of explanations have been proposed. The most prominent is the presence of unmeasured confounding in the observational studies. Although no such confounder has been identified, critics are so certain of this explanation as to dismiss any alternative as ‘special pleading’.4 However, as discussed in detail previously,5 the evidence that confounding completely explains this discrepancy is sparse. This explanation is testable—the finding of a confounding factor that would explain away the association would be compelling. However, none has emerged. The few observational studies that adjusted for socioeconomic factors yielded results similar to the unadjusted results, and several studies that found an inverse relation between HT and CHD were limited to participants with relatively homogeneous socioeconomic status. Also, adjustment for a large variety of measured confounders typically has only a modest impact on relative risk estimates. For example, Kim et al. report a crude odds ratio of 0.82 for the association between HT and CHD, and, after adjustment for 15 potential confounding factors, the odds ratio (OR) was not attenuated (OR = 0.76). A more telling argument against confounding as the complete explanation is the finding of nearly identical results for stroke in the observational Nurses' Health Study and in the Women's Health Initiative (WHI) trial.6 Since stroke and coronary disease share many risk factors (and, like CHD, stroke is also strongly linked to behavioural and societal influences), it is implausible that coronary disease findings could be entirely due to confounding while stroke findings were not so affected.
An alternative explanation, based on biology, proposes that post-menopausal hormones have different effects depending on a women's arterial health at the time hormone use is initiated.6 Thus, women who begin using hormones at menopause, when atheromatous plaques are uncommon, may benefit from HT, but not women who start at older ages, when such lesions are common and may be adversely affected by hormones. As described in detail in that commentary,6 such differential effects of HT are supported in animal studies, as well as limited studies in women. For example, among women with CHD, the flow-mediated vasodilatory response was similar in HT users and non-users, but, among women without CHD or major risk factors, HT users had a 40% better response than non-users. This is also consistent with the findings from the oestrogen–progestin arm of the WHI, which showed a distinct gradient by time since menopause.7 This hypothesis is also testable. Indeed, a year after that alternative explanation was advanced, the WHI oestrogen-only trial results were published that further supported such an age-at-initiation gradient, with a hazard ratio of 0.56 [95% confidence interval (95% CI) 0.30–1.03] for women ages 50–59 years,8 consistent with the findings of the observational studies. The vast majority of data from observational studies derive from women who started hormone use around the time of menopause, reflecting clinical practice. In contrast, the trials enrolled older women, mostly many years post-menopausal (those aged 50–59 years provided a small proportion of CHD events in the WHI trial).
Data regarding the relation of HT to CHD among women who initiated hormones at older ages are sparse in observational studies. In the Nurses' Study, we have observed no evidence for benefit among such women, but the CIs were broad owing to the small number of women who began hormones at older ages.9
In evaluating evidence regarding this ‘age-at-initiation’ hypothesis, it is important to distinguish between the age-at-initiation and the age of women during their hormone use. Women typically initiate HT near menopause, and some continue to take it for many years after menopause. Most studies find that such women have lower rates of CHD compared with never users. Few women first begin HT at older ages. Kim et al. state that they are testing the effect of older age-at-initiation; in fact they probably have examined mainly women who continued to take hormones into older ages.
Although the word ‘initiation’ is used throughout their paper, its unique meaning for this paper is provided in one brief sentence in the Methods section: ‘Age of HRT initiation was calculated using the first recorded instance of HRT use within the time frame of study observation.’ (italics added) and one sentence in the Discussion. Thus, ‘age at initiation’ is not used in its usual meaning of ‘age at beginning of use’, but simply first use within the observation period of this study. Probably the large majority of these women were not beginning HT use then, since most women first begin hormone use around the time of menopause. Interestingly, the two sentences explaining the use of ‘initiation’ in this paper were inserted only after I pointed this out to the authors and editors, after having scrutinized the data and deducing that these data could not be reflecting ‘initiation’ as that word is usually interpreted, since there were so many more women ‘initiating’ HT at older ages than could plausibly be expected given usual prescription patterns. While it would indeed be of great interest to assess the impact of age at initiation of HT on CHD risk, Kim et al. have not done so, despite their use of misleading terminology.
Abraham Lincoln once said, ‘How many legs does a dog have if you call the tail a leg? Four. Calling a tail a leg does not make it a leg.’
 |
References |
|---|
 Top References |
|---|
1 Grady D, Rubin SB, Pettiti DB et al. Hormone therapy to prevent disease and prolong life in postmenopausal women. Ann Intern Med 1992;117:1016–37.[ISI][Medline]
2 Kim J, Evans S, Smeeth L, Pocock S. Hormone replacement therapy and acute myocardial infarction: a large observational study exploring the influence of age. Int J Epidemiol 2006;35:731–38.
3 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.
4 Davey Smith G, Ebrahim S. Folate supplementation and cardiovascular disease. Lancet 2005;366:1679–81.[CrossRef][ISI][Medline]
5 Stampfer MJ. Hormones and heart disease: Do trials and observational studies address different questions? Int J Epidemiol 2004;33:1–2.
6 Grodstein F, Clarckson TB, Manson, JE. Understanding the divergent data on postmenopausal hormone therapy. N Engl J Med 2003;348:645–50.
7 Manson JE, Hsia J, Johnson KC et al. Estrogen plus progestin and the risk of coronary heart disease. N Engl J Med 2003;349:523–34.
8 Women's Health Initiative Steering Committee. Effects of conjugated equine estrogen in postmenopausal women with hysterectomy: the Women's Health Initiative randomized controlled trial. JAMA 2004;291:1701–12.
9 Grodstein F, Manson JE, Stampfer MJ. Hormone therapy and coronary heart disease: the role of time since menopause and age at hormone initiation. J Womens Health 2006;15:35–44.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Kim, S. Evans, L. Smeeth, and S. Pocock Response: Response to the Stampfer commentary Int. J. Epidemiol., June 1, 2006; 35(3): 740 - 740. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||