Measuring serum oestradiol in women with Alzheimer's disease: the importance of the sensitivity of the assay method.

OBJECTIVE Oestrogens could be protective against the development of Alzheimer's disease (AD) but reports on oestrogen levels in AD have been conflicting. DESIGN AND METHODS A meta-analysis using robust regression was carried out to assess whether the sensitivity of the assays of past studies had affected the reported level of total oestradiol. We had also measured total oestradiol in women with AD (n=66) and controls (n=62) not using hormone replacement therapy. We used two assays for total oestradiol to assess the difference between sensitive (radioimmunoassay with a specific rabbit antibody: 3 pmol/l) and relatively insensitive (immunoassay: 37 pmol/l) assays. RESULTS Meta-analysis using robust regression indicated that insensitive assays gave higher levels of total oestradiol when many samples fall below the level of sensitivity of the method. Earlier reports of low levels of total oestradiol in AD might be explained by this phenomenon, since total oestradiol levels (using the sensitive assay) in our controls were one third of those reported in the earlier studies. Using the sensitive assay we found that women with AD had significantly (P<0.01) higher levels (26+/-13 pmol/l) of total oestradiol than controls (21+/-13 pmol/l). Using the insensitive assay, there was no significant difference in the levels of total oestradiol. CONCLUSIONS The sensitivity of the assay determines the reported value of the oestradiol levels. Studies using a sensitive assay do not report significantly lower levels of total oestradiol in women with AD. This weighs against the hypothesis that low levels of total oestradiol are a risk factor for AD.