The Effect of Postmenopausal Estrogen Therapy on the Risk of Non-Insulin-Dependent Diabetes Mellitus
ABSTRACT & COMMENTARY
Synopsis: Estrogen therapy protection against diabetes may be due to confounding biases.
Source: Gabal LL, et al. Am J Public Health 1997; 87:443-445.
Gabal and colleagues from the university of California, San Diego, report on the incidence of non-insulin-dependent diabetes mellitus in the Rancho Bernardo cohort. The Rancho Bernardo Heart and Chronic Disease Study began in the early 1970s to assess heart disease risk factors in a retirement community. In this report, the onset of diabetes mellitus was assessed in 848 postmenopausal women followed for 10-15 years. There were 105 new cases of diabetes. There was no statistically significant evidence for a reduced risk of diabetes in women who were either past or current users of estrogen therapy. Gabal et al suggest that previous results showing a reduced risk of diabetes were due to confounding biases.
COMMENT BY LEON SPEROFF, MD
Two previous studies have found a reduced risk of non-insulin-dependent diabetes mellitus in women who are users of estrogen therapy (Am J Obstet Gynecol 1979;133:525; Ann Epidemiol 1992;2:665). Most notable was the Nurses’ Health Study, finding approximately a 20% reduced risk in this prospective cohort with large numbers. In the current study, although the results did not reach statistical significance, the lowest incidence of diabetes was found among continuous users of estrogen therapy, and the highest incidence was among women who had never used estrogen therapy. However, Gabal et al believe that this finding was due to the fact that the women who never used estrogen were older and more overweight, while the estrogen users were younger and thinner, drank more alcohol, and smoked more often. As a result, the current continuous users of estrogen in this group had lower fasting glucose levels right from the start. Thus, the effect of estrogen therapy on the risk of non-insulin-dependent diabetes is uncertain because these observational studies cannot free themselves from the confounding effects of various biases. Indeed, the answer will only come when the results from the Women’s Health Initiative, the ongoing, randomized clinical trial, become available.
Nevertheless, there is reason to believe that estrogen therapy may have a beneficial effect on the risk of non-insulin-dependent diabetes. Estrogen in the standard doses used for postmenopausal hormone therapy has a favorable effect on the hyperinsulinemia that is characteristic of women in the postmenopausal years. In addition, in women with diabetes mellitus, the administration of estrogen results in a decreasing glucose, improvement in hemoglobin A1C, as well as decreases in C-peptide and IGF1, all of which indicate improved carbohydrate metabolism (J Clin Endocrin Metab 1997;82:638). Such a beneficial effect on carbohydrate metabolism would have a major effect, contributing to protection against cardiovascular disease in postmenopausal women.
At this point, it seems to me that the evidence supports providing estrogen therapy to women who are known diabetics because they have much to gain in terms of protection against cardiovascular disease. In addition, there is biologic plausibility that, ultimately, the randomized clinical trial data will indicate a reduced risk of non-insulin-dependent diabetes.