The following information comes from a published letter on measuring estrogen: Shuer
Premarin is derived from horses, thus the “equine” below. When women take Premarin, they metabolize the “conjugated equine estrogen” it contains into the following forms of estrogen: estrone sulfate, equilin sulfate, 17[alpha]-dihydroequilin sulfate, equilin, equilenin, 17[beta]-dihydroequilin, or 17[beta]-dihydroequilenin.
None of these compounds is measured by standard lab tests for “estrogen” (radioimmunoassay). The form of estrogen produced by the human ovaries is primarily estradiol, specifically 17[beta]-estradiol. That can be measured by standard lab tests, and is reported as “estradiol”. But for a women taking Premarin, measuring an estradiol level is like measuring an imipramine level when a woman is taking Prozac!
Even ethinyl estradiol, which is the synthetic estrogen found in most oral contraceptives, is not picked up on lab tests for estradiol. In a woman taking an oral contraceptive (which blocks production of her own estrogen) you can’t tell how much estrogen she’s exposed to except by her symptoms, such as excessive uterine bleeding or breast tenderness — because the lab test won’t measure what she’s taking.
In the view of Dr. Shuer, this leads to a situation in which women can be exposed to high levels of estrogen and yet there’s no simple way to test for that. And therefore the studies of estrogen risk are being done when we can’t measure how much estrogen women are being exposed to! Thus it does seem odd to think that we’ve characterized the risks of estrogen, if we don’t know (e.g. in the Women’s Health Initiative study) how much they were getting.