Serum follicle stimulating hormone (FSH) level is measured on day three ( or 2 or 4) of the menstrual cycle. (First day of period flow is counted as day one. Spotting is not considered start of period.) If a lower value occurs from later testing, the highest value is considered the most predictive. FSH assays can differ somewhat so reference ranges as to what is normal, premenopausal or menopausal should be based on ranges provided by the laboratory doing the testing. Estradiol (E2) should also be measured as women who ovulate early may have elevated E2 levels above 80 pg/mL (due to early follicle recruitment, possibly due to a low serum inhibin B level) which will mask an elevated FSH level and give a false negative result.
High FSH strongly predicts poor IVF response in older women, less so in younger women. One study showed an elevated basal day-three FSH is correlated with diminished ovarian reserve in women aged over 35 years and is associated with poor pregnancy rates after treatment of ovulation induction(6% versus 42%).
This study by Gleicher et al assessed what normal values for FSH should be according to age. As you can see below the normal values are way below what we normally tell patients: less than 9 mIU/ml .
Since normal b-FSH levels rise with female age, these levels should represent a more accurate represent- tation of ovarian function than currently used universal cut-off levels for all ages. Women who exceed their age-specific cut off levels, should be suspected of demon- strating PREMATURE OVARIAN AGING and should, therefore, immediately, be directed towards further diagnostic evaluation.
AGE SPECIFIC b-FSH LEVELS
< 33 Years 33-37 Years 38-40 Years ≥ 41 Years
< 7.0 mIU/ml < 7.9 mIU/ml < 8.4 mIU/ml < 8.5 mIU/ml
abstract below (source fertility and sterility)
We evaluated a study group of 434 consecutive patients under age 41 years with baseline (b-) FSH levels of <12 mIU/ml (considered to represent “normal” ovarian function), who underwent ovarian stimulation for IVF with an ovarian stimulation protocol consisting of long GnRH-agonist or antagonist suppression and modal gonadotropin stimulation of 300IU of FSH/HMG per day. We assessed IVF cycle outcomes, including oocyte yields, based on age-specific b-FSH levels, defined as levels ≤ the 95% confidence interval of the mean (95% C.I.) for each age group. In the literature production of fewer than 5 oocytes in response to ovulation induction is considered to be evidence of ovarian resistance. We consider women under the age of 41 who produce fewer than 5 oocytes to have POA. Women with b-FSH levels above the 95% CI for their respective age groups were considered to be at increased risk of premature ovarian aging (POA). A logistic regression model for the presence of fewer than 5 oocytes at retrieval was performed using SPSS for Windows15.0. Continuous variables are presented as mean ±1 SE.