We were most interested by the recent article by Florio and colleagues (Vol 106, October 1999)1, describing the ontogeny of inhibins and activin A at parturition. The apparent shift in placental/fetal membrane secretion from inhibin A to activin A in association with term labour, suggesting a down regulation of inhibin α-subunit production in the tissues, has also been recently reported by Keelan and colleagues2, who quantified tissue content of these proteins at term and in association with labour. These two studies, together with that of Fowler et al.3 are consistent with increased bio-availability of activin at the time of labour and a possible role for activin in parturition. Whether such a role exists remains to be defined.
We were also intrigued by the observations by the authors that inhibin B was detectable in maternal serum and that inhibin A was detectable in umbilical cord serum. Previously, we have been unable to detect any inhibin B in maternal serum in early pregnancy4 and Fowler et al. observed only very low levels during the third trimester, ∼21 pg/mL, probably reflecting the known 10% cross-reactivity of inhibin A in the assay that has been common to all of these studies. Furthermore, ongoing work in our laboratory using immuno-affinity gel chromatography has also failed to detect inhibin B in pools of maternal serum collected throughout pregnancy including at term. The level of detection of the assay in our laboratory is 4 pg/mL. That Florio and colleagues found inhibin B levels of 70–200 pg/mL, which are much higher than would be accounted for by inhibin A cross-reactivity alone and well within detection of the assay in our hands, was therefore surprising. Similarly, we4 and others5 have previously reported inhibin levels in umbilical cord serum (both artery and vein) at term, observing inhibin B only in serum from male infants and no detectable inhibin A in any sample. For the purpose of this correspondence we have assayed inhibin A and inhibin B in 37 umbilical artery serum samples collected, within the last month, from uncomplicated term pregnancies following a normal vaginal delivery. Fourteen of these were from a female infant and 23 from a male. Inhibin A was undetectable in all 37 samples, with an assay detection limit of 15 pg/mL, and inhibin B was present only in serum from the 23 males (mean ± SEM = 160 ± 24 pg/mL), consistent with the levels we have reported previously4. We cannot explain the apparent discrepancies between out data, both previous and current, and those of Florio and colleagues1 but an understanding of these would seem important to the successful unravelling of the role(s) inhibins and activins may have in late pregnancy and parturition.