Age and Illness Cited for Women's Higher MI Death Rate

PARIS, Feb. 19 -- Women have a higher mortality rate than men after a heart attack largely because they are older and sicker when these events occur, according to investigators here.

It's not disparity of care but disparity of age and illness, reported Carine Milcent, Ph.D., and Brigitte Dormont, Ph.D., of the Paris-Jourdan Sciences Economiques research center, and colleagues.

Although women were indeed less likely to receive percutaneous coronary intervention than men among nearly 75,000 French women and men who had heart attacks in 1999, about 75% of the difference in in-hospital mortality rates between the sexes could be chalked up to women being older and having more co-morbidities.

Only about one-fourth to the lower use of percutaneous coronary intervention and stents contributed to women's higher mortality after an MI, the investigators wrote in the Feb. 19 Circulation, Journal of the American Heart Association, a special women's health issue.

The evidence suggested that while women would benefit from more frequent use of percutaneous coronary intervention, they appear to get less protection from revascularization procedures than men, the authors said.

"The explanation for the reduced protective effect against death afforded by percutaneous coronary intervention in women is unclear," they wrote. "It may be related to generally poorer outcomes of percutaneous coronary intervention in women (less benefit and higher complication rates), possibly because of smaller target-vessel size, increased vessel tortuosity, and other biological differences."

Those differences may include hormonal changes during and after menopause, differences in vascular structure, and, possibly, to sex-related differences in arterial size and remodeling, wrote R. David Anderson, M.D., and Carl J. Pepine, M.D., of the University of Florida in Gainesville, in an accompanying editorial.

"Women typically have smaller and less compliant conduit arteries than men. This is true even after adjustment for differences in height, weight, and blood pressure," they wrote. "Age-related stiffening of the aorta appears more prominent in diabetic women than men. Changes in arterial size have been documented to occur during pregnancy."

They noted that when a man receives a woman's heart in a transplant, the coronary arteries grow over time, whereas there is no similar change in artery size when a woman receives a heart graft from another woman. Conversely, male-to-female transsexuals (i.e., genetic men who take estrogen) have been demonstrated to have a reduction in the size of their brachial arteries, presumably related to hormonal change, the editorialists noted.

The investigators drew on a national database for information about all hospital admissions in 1999 with a discharge diagnosis of acute myocardial infarction. Their goal was to create microsimulation models exploring the variables that might account for observed excess mortality after acute MI among women, and to estimate outcomes if women had received the same care as men.

They conducted logistic regression analyses on mortality, controlling for age, comorbidities and coronary interventions.

They identified a total of 74,389 patients hospitalized for acute MI: 52,041 men (70%) and 22,348 women (30%). Women were on average 12 years older than men at the time of admission (75 versus 63, P<0.001), and women had more than twice the crude the in-hospital death rate as men (14.8% versus 6.1%; P<0.0001).

Men were more likely than women to undergo percutaneous coronary intervention (7.4% versus 4.8%, P<0.001), and of those who underwent a percutaneous intervention, men were also more likely to receive a stent (24.4% versus 14.2%, P<0.001).

When the investigators adjusted the analysis for age and for co-morbidities, the authors found that the death rate was still significantly higher among women (P<0.001), and that their excess adjusted absolute mortality was 1.95%.

The authors estimated that had the women had percutaneous coronary intervention at the same rate as men, their expected mortality rate would be 14.32%, compared with the actual rate of 14.78 -- a difference of 0.46%, or about one-fourth of the excess adjusted absolute mortality difference of 1.95%.

But when they plugged into their microsimulation model the assumption that women undergoing percutaneous coronary intervention at the same rate as men would have the same outcomes, they found that the expected mortality for women be 12.55%, suggesting that more than 90% of the difference in adjusted hospital mortality could be accounted for by a factor other than treatment bias.

"Thus, of the 8.64-point crude excess mortality in women with acute myocardial infarction, 6.69 is explained by the age structure of the population, 0.46 by the difference in procedure rates, and 1.77 by gender differences in the outcome of procedures and the impact of comorbidities," the authors wrote.

"The residual (-0.28) is related to differences in other characteristics (including unobservable characteristics). Thus, one quarter of the gender gap appears to be related to differential use of PCI plus stent between men and women."

The authors noted that men and women were equally likely to receive care in urban versus community hospitals, meaning that the gender differences could not be explained by differences in quality of care or institutional setting. Access to care was also not an issue, as France has universal and uniform access to care, regardless of income or employment status.

The investigators noted that the study was limited by a lack of information on racial/ethnic composition of the cohort, indications for percutaneous coronary intervention (elective versus emergency, for example), or factors such as presence of ST-segment elevation or patient eligibility for reperfusion therapy.

"Milcent et al have reaffirmed the increased mortality in women treated for acute myocardial infarction," Dr. Anderson and Dr. Pepine wrote in their editorial. "Other healthcare systems could benefit from such a large and complete data set, Despite the more advanced age and risk profile of women, after adjustment, the authors have shown that there is still disparity. Some of this discordance may be due to bias or underuse of aggressive therapy. It is likely that given the complexities in gender differences, biology also plays a role."