Hormone Therapy and Risk of Alzheimer Disease: Title and subTitle BreakA Critical Time

Interest in the effects of hormone therapy on health outcomes among postmenopausal women has perhaps never been greater. As recently reported,¹ the Women's Health Initiative (WHI) randomized clinical trial of postmenopausal hormone therapy terminated the combined estrogen-progestin arm (but not the estrogen-only arm) early after an average follow-up of 5.2 years (planned duration, 8.5 years).² Women who received estrogen plus progestin experienced a small but significant increase in the primary outcome, coronary heart disease; a nonsignificant trend toward an increase in the primary adverse outcome, invasive breast cancer; and a significant increase in a global index (which included the 2 primary outcomes plus stroke, pulmonary embolism, endometrial cancer, colorectal cancer, hip fracture, and death due to other causes) summarizing risk and benefit. However, the decision to stop this portion of the trial did not take into account several other important outcomes that may affect a woman's choice of whether to take hormone therapy, including potential effects on cognition. Two ancillary studies to the WHI randomized trial focus on risk of dementia and cognitive decline, but these outcomes are not considered in the WHI risk-benefit profile.

The potential role of hormone therapy in reducing the risk for dementia due to Alzheimer disease is a topic of great concern for women at midlife and beyond. Animal models indicate plausible mechanisms through which estrogen may protect against Alzheimer disease or other neurodegenerative disorders.³ These include neurotrophic and neuroprotective influences, enhancement of synaptic plasticity, effects on several neurotransmitter systems, reduction in β-amyloid formation from its precursor protein, and modulation of regional cerebral glucose metabolism. Human neuroimaging studies provide evidence that estrogen-containing hormone therapy influences the pattern of brain activation during memory processing, increases regional cerebral blood flow and glucose metabolism in temporal lobe structures that subserve memory, and modulates brain activity in specific brain regions affected in the early stages of Alzheimer disease.⁴

Despite the neurobiological rationale for a protective effect of estrogens on the risk for Alzheimer disease, results of observational studies of hormone therapy and Alzheimer disease have been inconclusive. Findings from a number of case-control studies and 2 prospective investigations suggest about a 30% reduced risk in women who report ever having received hormone therapy compared with nonusers.^{5 - 6}

In this issue of THE JOURNAL, Zandi and colleagues⁷ provide further observational evidence of a protective effect of hormone therapy on the incidence of Alzheimer disease in a large prospective population-based study in a delimited geographic area, Cache County, Utah. Consistent with the results of 2 prior prospective studies,^{8 - 9} the authors report that women who used hormone therapy had a reduced incidence of Alzheimer disease compared with nonusers over a 3-year follow-up interval (adjusted hazard ratio [HR], 0.59; 95% confidence interval [CI], 0.36-0.96). The apparent reduction in incidence of Alzheimer disease in women receiving treatment is less likely to be due to the "healthy user" bias, ie, the tendency for hormone users to be better educated and healthier overall. Specifically, reported use of multivitamins and calcium supplements was not associated with reduction in risk. Furthermore, as in other studies,^{8 ,10 - 11} longer duration of treatment was associated with greater risk reduction (adjusted HR, 0.41; 95% CI, 0.17-0.86, in women with >10 years of treatment). In addition, the results reported by Zandi et al⁷ indicate that former users of hormone therapy have a greater reduction in the incidence of Alzheimer disease than current users. Among current users, only long-time users (>10 years) appeared to benefit.

These findings raise important issues concerning current understanding of the way in which hormone therapy may potentially protect against Alzheimer disease, for the interpretation of existing studies, and for the design of new investigations. In the study by Zandi et al⁷ and other recent reports,^{12 - 14} women appear to be at increased risk for Alzheimer disease compared with men, particularly among the oldest-old. This approximately 2-fold excess risk was not observed among women reporting more than 10 years of hormone treatment. In contrast, current use of hormone therapy of less than 10 years' duration was not associated with protection against a diagnosis of Alzheimer disease. As the authors point out, the difference in findings for current vs former users of hormone therapy might indicate that treatment is ineffective during the latent preclinical stage of Alzheimer disease, which may extend a decade or more before the onset of diagnosable dementia.

The lack of an effect during this stage is consistent with the negative findings in recent treatment trials of patients who already have been diagnosed with the disease.^{15 - 17} It might also explain the contrast between the negative results of the 2 case-control studies based on prescription records within the preceding decade^{18 - 19} and the protective effects of treatment observed in the prospective studies that considered "ever-use" of hormone therapy.^{8 - 9} Because many women use hormone therapy for relatively brief durations around the menopause, the protective effect of ever-use therapy suggests the possibility of a critical period during the climacteric years, which are characterized by relatively rapid estrogen depletion.^{20 - 21} As Zandi et al⁷ note, their observation of a reduced risk in former but not current users, unless the latter were long-term users, is consistent with this hypothesis. Animal studies also provide preliminary support for a critical window of estrogen effects on cognitive function.²²

The possibilities of a critical period for hormone use in protection against Alzheimer disease and the lack of a benefit during the preclinical phase of this illness have significant implications for prevention trials. Several large randomized clinical trials of the effects of hormone therapy on risk for Alzheimer disease are under way—including the Women's Health Initiative Memory Study (WHIMS),²³ the Women's International Study of Long Duration Oestrogen After Menopause (WISDOM),²⁴ and the PREventing Postmenopausal Memory Loss and Alzheimer's With Replacement Estrogens (PREPARE) trial²⁵ —but all of these studies involve initiation of hormone treatment in women aged 65 years and older. It has been argued that estrogen-responsive neurons may show diminished response after a prolonged period of estrogen depletion.²⁶ Zandi et al⁷ appropriately caution that a beneficial effect of hormone therapy should not be ruled out if initial results of these trials are negative. It will be critical to examine prior use of hormone therapy, especially during the menopausal transition, as a modulator of later effects of treatment in these clinical trials. The possibility of a critical period during the climacteric also emphasizes the need for balance in considering data from observational studies and randomized clinical trials. If use of hormone therapy during the menopausal transition is required for later neuroprotection, a randomized clinical trial will be quite difficult because follow-up would be required over a period of decades.

The findings of Zandi et al⁷ are open to other interpretations as well. As in all observational studies, potential confounds may not be readily apparent. The absence of a protective association among short-term current users of hormone therapy could reflect the possibility that women with mild cognitive symptoms, who may be at increased risk for frank dementia,²⁷ are more likely to initiate hormone treatment. The relation between age and duration of use is not addressed directly by Zandi et al, but older women may be less likely to use hormone therapy for as long as younger women.²⁸ If so, the relation between timing of estrogen use could be confounded by duration of use, and the protective association of hormone therapy might then be less dependent on a critical, early window than on longer exposure. Another consideration is that long-term use might be related to other estrogen actions that putatively enhance or maintain cognitive function (eg, effects on synaptic plasticity, cerebral metabolism, or neurotransmitter modulation) rather than factors involved in Alzheimer disease pathogenesis per se (eg, retarding β-amyloid deposition).

Where does this new information and where do these new questions leave women and their physicians in making choices about whether and when to treat with hormone therapy? The current data are insufficient to recommend hormone therapy for prevention of Alzheimer disease.²⁹ However, many women will continue to consider short-term use of hormone therapy for treatment of menopausal symptoms, and the possibility of a critical period of use suggests that treatment during the climacteric might offer some protection against Alzheimer disease. The results reported by Zandi et al offer both hope for a possible neuroprotective effect of hormone therapy and frustration that it could be difficult to determine the optimal timing of treatment. Almost all hormone use in this cohort involved estrogen treatment unopposed by a progestin, which may offer the greatest potential for neuroprotection.

These new findings emphasize the need for continued research on the optimal regimen, dose, and timing of hormone therapy and other related compounds (such as selective estrogen receptor modulators) for possible neuroprotection against a diagnosis of Alzheimer disease years later. Although the combined estrogen-progestin arm of the WHI randomized trial has been terminated based on a specific risk-benefit profile for a specific therapeutic regimen, the risk-benefit profile may well change as other important outcomes and treatments are considered.