The numbers of 2-5 and 6-9 mm ovarian follicles are inversely correlated in both normal women and in polycystic ovary syndrome patients: what is the missing link?

STUDY QUESTION: Is the negative correlation between the numbers of 2-5 and 6-9 mm follicles influenced by ovarian and/or metabolic parameter(s) in young control women and in patients with polycystic ovarian syndrome (PCOS)? SUMMARY ANSWER: Our study confirmed that the negative correlation between numbers of follicles sized 2-5 and 6-9 mm was stronger in PCOS than in young control women and was not linked to any ovarian or metabolic parameter. WHAT IS KNOWN ALREADY: Previous reports described a direct negative correlation between the number of small antral follicles (2-5 mm) and large antral follicle (6-9 mm) during the early follicular phase (cycle Days 2-5) in normal and PCOS women. Numerous factors, that could be either intrinsic to the ovary or secondary to metabolic influence and/or gonadotropin regulation, might account for this. STUDY DESIGN, SIZE, DURATION: Six hundred and thirty-nine patients with PCOS according to Rotterdam Criteria and 157 control women were recruited in this retrospective cross-sectional study from January 2009 to January 2016. PARTICIPANTS/MATERIALS, SETTING, METHODS: Data were obtained from a database of clinical, hormonal and ultrasound (U/S) features recorded consecutively in a single reproductive medicine centre. Univariate correlations between the various parameters were analysed by the Spearman's correlation test. All variables significantly related to the 2-5 and/or 6-9 mm follicle numbers were included in a principal component analysis (PCA) in order to structure the data and to obtain collections of uncorrelated variables, called principal components (PC), which are linear combinations of the original variables. MAIN RESULTS AND THE ROLE OF CHANCE: By univariate analysis, the 2-5 and 6-9 mm follicle numbers were strongly but negatively correlated in both populations. Many other variables were correlated to the 2-5 and/or 6-9 mm follicle numbers and to each other. By PCA, these relationships were gathered into four independent PCs in each population. In both groups, the 2-5 and 6-9 mm follicle numbers correlated strongly and inversely to a specific PC. Among the other variables tested, only serum oestradiol level correlated weakly to this PC in the control group. Two other uncorrelated PCs gathered relationships between variables linked to the metabolic status and the gonadotropin regulation both in control and PCOS women. Lastly, a fourth PC included relationships which linked to ovarian ageing in controls and to follicle dysregulation in patients with PCOS. LIMITATIONS, REASONS FOR CAUTION: Our controls did not represent the general population since they were recruited in an ART centre; we used a modified Rotterdam classification for PCOS using follicle count and/or serum AMH level with in-house thresholds to define the follicle excess; the AMH assay used is no longer commercially available. WIDER IMPLICATIONS OF THE FINDINGS: Factor(s) regulating specifically the equilibrium between the 2-5 and 6-9 mm follicle numbers still need(s) to be identified. More attention should be paid to the oocyte. STUDY FUNDING/COMPETING INTEREST(S): None.