POSTMENOPAUSAL HORMONE THERAPY AND REGIONAL BRAIN VOLUMES: THE WHIMS-MRI STUDY

AND REGIONAL BRAIN VOLUMES: THE WHIMS-MRI STUDY To the Editor: We read the article by Resnick et al.1 regarding findings from the Women’s Health Initiative Memory (WHIMS)–MRI Study on hormone therapy (HT) and brain volumes on MRI. The investigators found slightly lower hippocampal volumes on MRI after a 3-year follow-up in women assigned to HT compared to placebo. This confirms our previous observation in the Rotterdam Scan Study, a population-based MRI study in nondemented elderly where we found no beneficial effect of high endogenous plasma estradiol levels on manually measured hippocampal volumes on MRI in 210 randomly selected women.2 If anything, we found that higher estradiol levels were associated with lower hippocampal volumes. For a portion of this cohort (n 94), follow-up MRIs are now available and were made an average of 3 years after baseline. Using an automatic segmentation procedure, baseline and follow-up hippocampal volumes were obtained in 94 women not using HT.3 Baseline total estradiol levels were not significantly associated with baseline hippocampal volumes on MRI. However, women in the lowest tertile of total estradiol at baseline had the largest decline in hippocampal volume. Estradiol levels at baseline were not associated with refusing the second MRI or risk of dying before the second MRI. The WHIMS investigators previously reported that the use of HT increased the risk of dementia and suggest in their current report that this effect is mediated by increased hippocampal atrophy.4 HT leads to much higher estradiol levels compared to the range of endogenous levels, for which we found an increased hippocampal atrophy rate in women with the lowest levels of estradiol. Combined, these findings suggest that either too low or very high level of estrogens could be detrimental to the hippocampus. An animal study has also indicated that estrogens could be neuroprotective except when evident in high doses.5