MHT and Alzheimer Risk: New Evidence for a Critical Window, with Roksana Karim, MD, PhD

A secondary analysis of the Early vs Late Intervention Trial with Estradiol (ELITE) found that menopausal hormone therapy (MHT) significantly accelerated the decline in amyloid beta-40 levels in postmenopausal women, with numerically greater effects on other Alzheimer disease biomarkers in women who started treatment within 6 years of menopause.

The findings were presented at The Menopause Society (TMS) 2025 Annual Meeting, October 21-25, by Roksana Karim, MD, PhD, professor of clinical population and public health sciences at the Keck School of Medicine of USC, in Los Angele, CA.

Roksana Karim, MD, PhD

Courtesy of Keck School of Medicine of USC

Karim and colleagues reported that, notably, MHT showed no effect on biomarkers in women who were 10 or more years past menopause. The researchers concluded that these findings partially support the critical window hypothesis—that hormone therapy's neuroprotective effects may depend on early initiation after menopause—and underscore the need for larger confirmatory studies.

Patient Care© met with Karim during TMS to discuss the findings of the analysis in greater depth. The following transcript of the interview has been lightly edited for length, style, and flow.

Patient Care: What prompted you and your colleagues to look at Alzheimer disease biomarkers in the Early versus Late Intervention Trial with Estradiol, or ELITE trial?

Roksana Karim, MD, PhD: The ELITE trial was completed in 2012, and the primary results were published in 2016, showing some promising results for initiating hormone therapy early after menopause on cardiovascular health. We saw significant benefit in the primary preventive marker of cardiovascular disease, atherosclerosis, when starting hormone therapy sooner in postmenopausal women.

As a follow-up, our secondary outcome was cognition. However, we did not see any improvement in cognition for early initiation of hormone therapy in those same women who saw cardiovascular benefit. This was puzzling to us. With the emergence of plasma biomarkers of Alzheimer disease, which are now more reliable and can be measured through blood rather than requiring the invasive process of collecting cerebrospinal fluid from the spinal cord, we saw an opportunity.

The FDA is even approving some of these diagnostic tests for early screening purposes. This development ignited our interest in looking at these biomarkers. Fortunately, we had stored samples from the ELITE participants longitudinally at every visit. We thought this would be the most impactful continuation we could do with those samples.

PC: Could you briefly explain what amyloid β-40 and β-42 tell us about Alzheimer disease risk, and why the ratio between them matters?

Karim: Both of these peptides are secreted in both plasma and cerebrospinal fluid. In a normal, healthy person, these peptides are supposed to be cleared out of the system. However, when they're not cleared properly, they tend to deposit in brain tissue. These deposits form amyloid plaques, which are the hallmark of Alzheimer disease.

Alzheimer disease has a long pathogenesis. Early on, people start showing symptoms of mild cognitive impairment, and that progresses over time to full-blown Alzheimer disease. Amyloid beta-42 (Aβ-42) is more of a villain than amyloid Aβ-40 because in about two-thirds of Alzheimer cases, Aβ-42 is the component that starts depositing in the brain and producing the plaques. When this happens, the concentration in plasma and cerebrospinal fluid decreases because the peptide is being deposited in brain tissue, where it starts harming memory.

Amyloid Aβ-40 is also deposited in amyloid plaques, but not to the same degree as Aβ-42, so its concentration remains higher in cerebrospinal fluid and plasma. Any indication that β-42 levels are going down suggests that it's being deposited in the brain.

PC: Your analysis found that menopausal hormone therapy significantly accelerated the decrease in amyloid Aβ-40. What does this finding mean clinically?

Karim: As I mentioned, Aβ-40 is the lesser of the 2 evils, but it's still harmful. When Aβ-40 concentration is higher, it's going to deposit once β-42 has started forming plaques, and it will worsen the plaques. So higher Aβ-40 concentration is harmful for us. If hormone therapy can reduce this concentration, that's somewhat good news.

PC: You saw these effects on biomarkers in early postmenopausal women, but not so much in women who were 10 years out from menopause. What's the significance of that finding, and does it help explain the critical window hypothesis?

Karim: I think it does. Estrogen receptor sensitivity and the ability to clear these harmful peptide components from the brain are functions of a healthy hormonal environment in the body and healthy brain structure. With menopause, this starts declining. The tissues take time to become nonresponsive to natural hormones or supplemental hormones, which is why the critical window hypothesis states that the sooner a woman is exposed to supplemental estradiol, the better the chance of responding to its molecular activity. And we have seen in some trials that starting hormone therapy later would increase the probability of an increased risk of cognitive decline.

PC: Most of the findings in early postmenopausal women showed numeric trends, but they didn't quite reach statistical significance. Should that be interpreted as a lack of power, or are the effects genuinely modest?

Karim: Remember that these results are very early. We really need to look more carefully at the layers of these analyses. There are many factors at play. For instance, in addition to the time-since-menopause variable, we also have the ApoE4 genotype layer, which increasingly reduces our power to see the actual effect if it's there.

I would suggest that we look at these results with hope, but it would be premature to jump to conclusions from this one finding. We need to give it time for others to reproduce these results. As the ELITE group, we also need to dig deeper into these findings.

PC: So it's premature to speak to postmenopausal or perimenopausal women about the critical window and these findings? These are just very tantalizing first results?

Karim: That's for sure, with regards to cognition and the results we presented at The Menopause Society meeting yesterday. However, for the cardiovascular findings, I think that's very clear. The way the ELITE trial was designed and the formulation of estradiol therapy, everything worked out in favor of initiating hormone therapy soon after menopause in that group.

PC: What would be your take-home message for primary care physicians who are caring for women in this stage of life?

Karim: I would suggest that they keep an eye out for more research on this issue. They could also initiate clinical trials looking at this question themselves. The more data we have, the better it is for us to establish these findings quickly.