Women's health initiative: no decline in dementia risk with estrogen use; Tied to insignificant increases in probable dementia, mild cognitive impairment.
The study included a subgroup of women from the estrogen alone and combined estrogen plus progestin arms of the Women's Health Initiative (WHI). Unopposed estrogen therapy was associated with insignificant increases in both probable dementia and mild cognitive impairment, compared with placebo.
When the data for estrogen alone and estrogen plus progestin were pooled, the hazard ratio for the composite end point of probable dementia or mild cognitive impairment increased significantly.
"This likely will not change any of the guidelines--this is just more information on the lack of utility of hormone therapy for the prevention of the chronic diseases of aging." Dr. Jan Shifren, director of the menopause program of the Vincent Obstetrics and Gynecology Service at the Massachusetts General Hospital in Boston, commented in an interview.
Current recommendations from the American College of Obstetricians and Gynecologists and the North American Menopause Society advise against the use of hormone therapy (HT) for anything other than the relief of menopausal vasomotor symptoms. The new data from the Women's Health Initiative Memory Study (WHIMS) are consistent with these recommendations.
"Use of hormone therapy to prevent dementia or MCI [mild cognitive impairment] in women 65 years of age or older is not recommended," reported Sally A. Shumaker, Ph.D., of Wake Forest University. Winston-Salem, N.C., and her colleagues (JAMA 291:2947-58, 2004).
"It's very important to remember that these analyses are in older women and are clearly not addressing the effects of hormone therapy in healthy 50-year-olds who are taking hormones for hot flashes," Dr. Shifren stressed.
"The study was designed to answer whether hormone therapy prevents dementia in high-risk women--in other words, older women--and the answer is, it does not," she said.
In a written statement, Dr. Gary L. Stiles, executive vice president and chief medical officer for Wyeth Pharmaceuticals, which markets various HT formulations including the ones used in the study, noted: "Age is the predominant difference between the women who use hormone therapy in current clinical practice and the women evaluated in WHIMS. The women in the WHIMS trial were on average nearly 20 years older than the typical newly menopausal woman. In addition, it is well established that risk of dementia increases dramatically with age, with or without hormone use."
After the estrogen plus progestin arm of the WHI was halted in 2002, WHIMS data published a year later showed an increased risk of dementia and no protection against MCI associated with estrogen plus progestin (JAMA 289:2651-62, 2003).
However, in keeping with many of the other risks reported with estrogen plus progestin therapy, there was speculation that the progestin component was responsible for the increased risk.
Dr. Shifren says the new WHIMS data are important because they invalidate this theory.
"It pools both the estrogen only and the estrogen plus progestin data, and both groups are completely consistent--showing the same risk for probable dementia. This shows that progestin is not the culprit," she said.
In the pooled analysis, the hazard ratio was significant for probable dementia (1.76) but insignificant for MCI (1.25), compared with placebo. The hazard ratio for the combination of these two end points (1.41) was statistically significant.
In the estrogen only trial, the hazard ratios were insignificant for probable dementia (1.49) and mild cognitive impairment (1.34), but significant for the composite end point (1.38).
In subgroup analyses, estrogen users in the estrogen alone trial had higher hazard ratios for probable dementia, compared with those receiving placebo, if they had diabetes (9.03 vs. 1.10) or hypertension (2.40 vs. 0.98).
The data were taken from a subset of WHI participants who were randomized to either 0.625 mg/day of conjugated equine estrogens (CEE) (2,947 women), or 2.5 mg/day of medroxyprogesterone acetate (MPA) plus 0.625 mg/day of CEE (4,532 women) during a follow-up period of 4-5 years.
In a separate WHIMS analysis, researchers evaluated the effect of HT on Global Cognitive Function in 1,387 women taking CEE and 2,131 women taking CEE/MPA, compared with 1,421 women taking placebo. All women were over age 65 (JAMA 291:2959-68, 2004).
Those receiving estrogen alone therapy demonstrated a slight but significant decrease in cognitive function--a mean decrease of 0.26 units on the Modified Mini-Mental State Examination (3MSE)--compared with women receiving placebo. This difference emerged 1-2 years after the initiation of therapy and persisted throughout the trial.
The results remained similar when the CEE results were pooled with the CEE/MPA results, reported Mark A. Espeland, Ph.D., of Wake Forest University, Winston-Salem, N.C., and his associates.
Overall, HT had a relatively greater adverse effect on women who had the lowest baseline levels of cognitive function, according to the investigators.
These negative effects support the conclusion that hormone therapy "initiated in the late postmenopausal period does not improve global cognitive function and may even adversely affect this outcome," the authors said.
In an editorial accompanying the analyses, Dr. Lon S. Schneider commented that the question remains whether early, short-term use of estrogen at the onset of menopause might offer more protection against late-life dementia (JAMA 291:3005-07, 2004).
Still, 45% of participants who reported earlier hormone use showed no benefits, suggesting that this intervention may not be useful either, said Dr. Schneider of the University of Southern California, Los Angeles.
BY KATE JOHNSON
|Printer friendly Cite/link Email Feedback|
|Publication:||OB GYN News|
|Date:||Aug 1, 2004|
|Previous Article:||Election year issue: doctors push for liability caps.|
|Next Article:||WHI database: increased ca rates seen in women with infertility Hx; Ovarian, endometrial, and breast cancer.|