Short answer: In women who started at 65 or older, it did in one large trial. In the WHI Memory Study, probable dementia ran 45 versus 22 cases per 10,000 women-years on estrogen plus progestin, and 37 versus 25 on estrogen alone, a rise that was not statistically significant. Trials in women who started in their early 50s found no overall effect on thinking. In November 2025 the FDA said it would take dementia out of the boxed warning.
The Women's Health Initiative Memory Study (WHIMS) enrolled women aged 65 to 79 from the WHI hormone trials, all free of probable dementia at the start. Both trials used daily oral conjugated estrogens (CE) 0.625 mg; the combined trial added medroxyprogesterone (MPA) 2.5 mg. See Women's Health Initiative.
| Trial | Women | Cases (hormone vs placebo) | Rate per 10,000 women-years | Hazard ratio |
|---|---|---|---|---|
| Estrogen plus progestin (Shumaker 2003) | 4,532, mean follow-up about 4 years | 40 vs 21 | 45 vs 22 | 2.05 (statistically significant) |
| Estrogen alone (Shumaker 2004) | 2,947 after hysterectomy, mean 5.2 years | 28 vs 19 | 37 vs 25 | 1.49 (not statistically significant) |
Pooled as the study planned, the hazard ratio was 1.76. Alzheimer disease was the most common type of dementia in both groups. Hormones also did not prevent mild cognitive impairment. The authors concluded that hormone therapy is not recommended to prevent dementia or memory decline in women 65 or older.
Two trials looked at younger women. WHIMS-Y followed 1,326 women who had joined the WHI hormone trials at ages 50 to 55 and taken CE, with or without MPA, for about 7 years. Tested on average 7.2 years after the trials ended, their overall thinking scores were similar to placebo. The authors found no overall lasting benefit or risk.
KEEPS-Cog randomized 693 women, average age 52.6 and 1.4 years past their last period, to oral CE 0.45 mg, an estradiol patch of 50 mcg a day (each with micronized progesterone 200 mg for 12 days a month) or placebo, for up to 4 years. Neither hormone group did better on memory or thinking tests. Women on oral CE reported fewer depression and anxiety symptoms; the patch group was similar to placebo. The authors noted the results apply to recently menopausal women with low heart risk and to 4 years of use.
The estrogen labels we read, including Climara (March 2026), Premarin (June 2026) and Prempro (April 2026), still carry a boxed warning on probable dementia in women 65 and older, say estrogen should not be used to prevent dementia, and say it is unknown whether the finding applies to younger women (section 5.3). See boxed warning.
On November 10, 2025, the FDA announced it was starting to remove the broad boxed warnings from menopause hormone products and working with companies to remove references to heart disease, breast cancer and probable dementia. It said the endometrial cancer boxed warning stays for systemic estrogen-alone products, and that the labeled recommendation will be to start systemic hormones within 10 years of menopause or before age 60. Label text changes as each company updates it.
Age at the start, years since the last period, the type of estrogen and progestogen and the reason for treatment all shape the discussion. Women worried about memory can ask what the trials do and do not show for their age. Dose Track keeps the product, dose and start date in one log, which is the history a prescriber will ask about.
Read more on Dose Track: Women's Health Initiative · Menopause · Premarin side effects · Hazard ratio