Value of basal FSH concentrations: prognostic implications for pregnancy outcome

Allison Styne-Gross1,2,3, Karen Elkind-Hirsch2 and Richard T. Scott2

1 Division of Reproductive Endocrinology and Infertility, Emory University School of Medicine, Atlanta, GA, USA 30322 and 2 Reproductive Medicine Associates of New Jersey, Morristown, NJ, USA 07962

3 To whom correspondence should be addressed. Email: astyne{at}aol.com

Sir,

We have read with interest the recent paper by van Montfrans et al. (2004)Go. In this study, the authors prospectively evaluated 129 women with no history of subfertility who were planning to attempt conception in the near future, and measured their basal FSH levels during three menstrual cycles. Following evaluation of the data, the authors concluded that basal FSH concentrations as a marker of ovarian ageing were not related to the incidence of early pregnancy loss (EPL).

We do not question the premise of the study since there are data in the literature associating diminished ovarian reserve with aneuploidy and the subsequent risk for pregnancy loss (Levi et al., 2001Go). However, there are some very significant issues with the methodology used in this study which have a dramatic impact on the interpretation of these data. It would be most concerning if a practicing clinician were to read this manuscript and conclude that ovarian ageing as evidenced by basal FSH levels does not impart a higher pregnancy loss risk.

Most critical to the interpretation of these types of data is how an abnormal result is defined. The accuracy of the diagnosis of ovarian ageing (commonly termed diminished ovarian reserve) is principally dependent on what FSH level is selected to distinguish normal from abnormal. The authors use a value of 12.5 IU/l and refer to a prior publication from their group (van Montfrans et al., 2000Go). A review of that publication demonstrates that much of the analysis was done using a level of 10 IU/l and that the pregnancy and delivery rate above that level were substantial. No threshold analysis was done. Foregoing the inconsistency in definitions by the authors, it is evident the threshold between normal and abnormal—typically selected at a level where the residual pregnancy rate is between 1 and 2%—should in fact be much higher. Based on their prior study, a threshold for an elevated basal FSH level would certainly be >20 IU/l in their laboratory. It is entirely possible and perhaps even likely that not a single patient in this study had an FSH level consistent with advanced ovarian age.

Even if the arbitrary definition of 12.5 IU/l were accepted, there were only five patients in the study whose levels exceeded that threshold. That number is certainly inadequate to accurately characterize the pregnancy and loss rates for this population of women. If as the title states the authors are evaluating the impact of ovarian ageing on early pregnancy loss, the power analyses should have been based on the study population (those with a study diagnosis of ovarian ageing), which means a sample size of five women at most. That being the case, the sample size is woefully inadequate.

If the authors were to argue that the entire population should be studied, then the focus of the study would shift from evaluating the impact of ovarian ageing on EPL, and would become an evaluation of whether or not variability in basal FSH levels within the normal range has predictive value for pregnancy outcome. The data as presented here clearly demonstrate that it does not. Having said that, one might question why anyone would suspect that it would. Variability within the normal range for endocrine hormones rarely, if ever, predicts changes in clinical performance or outcome. A suitable analogy might be found when evaluating the clinical symptoms associated with hyperthyroidism. For example, applying the methodology used in this study to evaluate the relationship between hyperthyroidism (as evidenced by low TSH levels) and tachycardia, the authors would bring in 129 women and measure their serum TSH levels and heart rates. Almost all of them would have TSH levels in the normal range. It is extremely probable that no relationship with tachycardia and TSH levels would be identified. Would it then be legitimate to conclude that hyperthyroidism as diagnosed by TSH levels has no relationship with tachycardia? While seeming silly, it is an exactly analogous situation. Variability in basal FSH levels within the normal range has never been demonstrated to be predictive of outcome.

Perhaps this paper might more correctly be entitled ‘Variations in basal FSH within the normal range does not predict early pregnancy loss.’

References

van Montfrans JM, van Hooff MHA, Huirne JA, Tanahatoe SJ, Sadrezadeh S, Martens F, van Vugt JMG and Lambalk CB (2004) Basal FSH concentrations as a marker of ovarian ageing are not related to pregnancy outcome in a general population of women over 30 years. Hum Reprod 19, 430–434.[Abstract/Free Full Text]

Levi AJ, Raynault MF, Bergh PA, Drews MR, Miller BT and Scott RT Jr (2001) Reproductive outcome in patients with diminished ovarian reserve. Fertil Steril 76, 666–669.[CrossRef][ISI][Medline]

van Montfrans JM, Hoek A, van Hooff MH, de Koning CH, Tonch N and Lambalk C (2000) Predictive value of basal follicle-stimulating hormone concentrations in a general subfertility population. Fertil Steril 74, 97–103.[ISI][Medline]





This Article
Extract
Full Text (PDF )
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Request Permissions
Google Scholar
Articles by Styne-Gross, A.
Articles by Scott, R. T.
PubMed
PubMed Citation
Articles by Styne-Gross, A.
Articles by Scott, R. T.