Little information is available about the risk of myocardial infarction (MI) in women with diabetes who also are using hormone therapy (HT). Ferrara and colleagues conducted a three-year observational study of a cohort of 25,000 women 50 years and older who had diabetes and were using different hormone regimens.

Women selected from the Northern California Kaiser Permanente Diabetes Registry were asked to complete a survey on risk factors for heart disease. Death-certificate information and the health plan’s electronic files were used to follow the study participants from January 1995 to one of the following end points: acute MI (1,199 participants), death not related to MI (2,010 participants), departure from the health plan (1,975 participants), or December 1998 (19,916 participants). The health plan’s computerized pharmacy system was used to track HT prescriptions, and the Cox proportional hazards model was used to determine whether MI risk was associated with HT use.

The risk of acute MI was 16 percent lower in women with diabetes who were taking HT and had no recent history of MI than in women with diabetes who had no recent history of MI and were not taking HT. Low or medium estrogen dosages were associated with a lower risk of MI, while high dosages (more than 0.625 mg of oral estrogens or 0.1 mg of estradiol) were not related to a lower risk. Among the women with a recent history of MI, HT was associated with an 80 percent higher rate of recurrent MI, especially in the first year of HT use.

The authors acknowledge that the results in women with diabetes but no recent MI were similar to the results from observational studies in women without diabetes, and that the results contradicted those from the Women’s Health Initiative, which reported a 30 percent increased risk of heart disease in women using HT. A possible explanation for the discrepancy is the lack of control and the potential bias inherent in observational studies. Even though the investigators accounted for various risk factors for heart disease, they might have missed other confounding variables showing HT as a marker for a better heart disease risk profile. The results in patients with diabetes and a recent history of MI are consistent with those from the Heart and Estrogen/Progestin Replacement Study.

The authors conclude that their results do not prove causality and that there is a need for more randomized clinical trials of HT in women who have diabetes.