Studies have shown some benefit to the use of estrogen in the treatment of Alzheimer's disease. However, these studies have been short term, have enrolled small numbers of patients and have not always used a standardized definition for the disease. Mulnard and associates conducted a randomized controlled trial to see if there were beneficial effects (global, cognitive or functional) related to the use of estrogen in elderly women with Alzheimer's disease.

Patients were enrolled from sites participating in the Alzheimer's Disease Cooperative Study. Women older than 60 years who had undergone a hysterectomy were included if they had a diagnosis of probable Alzheimer's disease (as defined by criteria of the National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer's Disease and Related Disorders Association). The patients' Mini-Mental State Examination (MMSE) scores were between 12 and 28 to restrict the study to participants who had mild to moderate dementia. Patients were excluded if they had a major depression or abnormal gynecologic, breast or mammography results. Patients were also excluded if they had experienced a recent myocardial infarction or thromboembolic disease or were taking certain disallowed medications, such as anticonvulsants, anticoagulants or narcotics. Continued use of neuroleptics, antidepressants, anxiolytics and sedative-hypnotics was allowed if the dosages were stable. Patients were randomized to receive placebo or estrogen in a dosage of 0.625 mg daily or 0.625 mg twice daily. After 12 months of placebo or estrogen, all patients entered a placebo wash-out phase for three months. A caregiver was required to give all of the study medications. Evaluations at screening, baseline, two, six, 12 and 15 months included cognitive measures and measures of mood, language and motor behavior.

A total of 120 women were enrolled in the study. All groups were similar at baseline. Efficacy was evaluated by comparing the placebo group (39 patients) with the combined estrogen groups (81 patients). There was no difference in the percentage of patients whose status worsened when the placebo group was compared with the estrogen group. The placebo group's results on the Clinical Dementia Rating scale was significantly better than the estrogen group's results. There were no significant differences between the estrogen and placebo groups in terms of evaluations of mood, memory, attention or activities of daily living.

A similar lack of effect was found when the estrogen groups were analyzed by dosage and compared with the placebo group. However, low-dose estrogen was found, as in other studies, to be related to an improvement in MMSE score at the two-month evaluation; this effect did not persist with continued treatment.

The authors conclude that estrogen does not improve outcomes (cognitive or functional) in women with mild to moderate Alzheimer's disease who receive this treatment for one year. A slight benefit seen at two months with 0.625 mg per day of estrogen does not persist if treatment continues. Estrogen should not be recommended as a treatment for Alzheimer's disease. In a related editorial, Shaywitz and Shaywitz concur that the evidence does not support treating Alzheimer's disease with estrogen, although they note that Mulnard's study was conducted in a specific set of patients (elderly women with mild to moderate severity of disease). Applying experimental findings clinically must be done cautiously and only after appropriately designed trials are completed. It is still possible that estrogen may play a role in preventing or delaying the onset of Alzheimer's disease.

editor's note: This study points out why randomized controlled trials are essential before clinical applications can be made. Previous prospective cohort studies have shown, some even fairly convincingly, that estrogen can reduce the risk of Alzheimer's disease or its symptoms, but this randomized controlled trial clearly does not support those earlier findings. This study found no benefit in those patients receiving estrogen; in fact, on some measures, placebo treatment seemed to show some (nonsignificant) trends toward benefit. At this point, estrogen replacement therapy may be indicated for other reasons, but not for the treatment of Alzheimer's disease.—g.b.h.