Genetic variations altering FSH action affect circulating hormone levels as well as follicle growth in healthy peripubertal girls.

STUDY QUESTION: Do variants of the genes encoding follicle stimulating hormone (FSH) beta subunit (B) and FSH receptor (R) impact circulating reproductive hormone levels and ovarian follicle maturation in healthy peripubertal girls?

SUMMARY ANSWER: FSHB and FSHR genetic variants exert, alone or their combination, distinct effects on reproductive hormone levels as well as ovarian follicle maturation in healthy peripubertal girls.

WHAT IS KNOWN ALREADY: FSHB and FSHR genetic variants impact reproductive hormone levels as well as associated pathologies in women. While FSHR c. 2039A>G is known to alter gonadotrophin levels in women, FSHR c.-29G>A has not yet been shown to exert effect and there are conflicting results concerning FSHB c.-211G>T.

STUDY DESIGN, SIZE, DURATION: This population-based study included 633 girls recruited as part of two cohorts, the COPENHAGEN Puberty Study (2006-2014, a cross-sectional and ongoing longitudinal study) and the Copenhagen Mother-Child Cohort (1997-2002, including transabdominal ultrasound (TAUS) of the ovaries in a subset of 91 peripubertal girls).

PARTICIPANTS/MATERIALS, SETTING, METHODS: Clinical examinations, including pubertal breast stage (Tanner's classification B1-B5) were performed. Circulating levels of FSH, luteinizing hormone (LH), estradiol, anti-Mullerian hormone (AMH) and inhibin-B were assessed by immunoassays. In a subset of the girls (n = 91), ovarian volume and the number/size of antral follicles were assessed by TAUS. Genotypes were determined by competitive PCR.

MAIN RESULTS AND THE ROLE OF CHANCE: FSHR c.2039A>G minor alleles were positively associated with serum FSH (β = 0.08, P = 0.004), LH (β = 0.06, P = 0.012) and estradiol (β = 0.06, P = 0.017) (adjusted for Tanner stages). In a combined model, FSHR c.-29G>A and FSHR c.2039A>G alleles were positively associated with FSH levels in early-pubertal girls (B2 + B3, n = 327, r = 0.1, P = 0.02) and in young adolescents (B4 + B5, n = 149, r = 0.2, P = 0.01). Serum AMH and inhibin B levels were not significantly influenced by the single nucleotide polymorphisms (SNPs). Single SNPs were not associated with follicles counts, however, a cumulative minor allele count (FSHB c.-211 G>T and FSHR c.-29G>A) was negatively associated with the number of large follicles (≥5 mm) (n = 91, P = 0.04) (adjusted for Tanner stages).

LIMITATIONS, REASONS FOR CAUTION: Since we studied girls and young adolescents during pubertal transition, our study may not be fully comparable with previous studies on FSHB and FSHR variants in adult women. The group of young adolescents (Tanner B4 + B5) reflects the endocrine situation in adult women best, however, the group is not large enough to contribute substantially to the conflicting results concerning the influence of FSHB c.-211G>T in adult women. Furthermore, we have no information about the exact day of the menstrual cycle in the subgroup of girls with menarche.

WIDER IMPLICATIONS OF THE FINDINGS: The sex-specific interaction of FSHB and FSHR genetic variants and physiological as well as pathological conditions is being increasingly elucidated. The variant triplet set might serve as diagnostic and pharmacogenetic marker. For the first time, we show an additional effect of FSHR c.-29G>A on serum FSH levels in healthy girls. Moreover, morphological data suggest impaired FSH-induced maturation of ovarian follicles in minor allele carriers of FSHB c.-211G>T and FSHR c.-29G>A. This may explain previous findings of delayed pubertal onset in these girls.

STUDY FUNDING/COMPETING INTERESTS: Funding was provided by the Danish Agency for Science, Technology and Innovation (09-067180), Danish Ministry of the Environment, CeHoS (MST-621-00065), Capital Region of Denmark (December 2011), Ministry of Higher Education and Science (DFF-1331-00113) and EDMaRC (Danish Ministry of Health). A.S.B. was funded from December 2015 by ReproUnion (EU Interreg Öresund-Kattegat-Skagerrak). The authors declare no conflict of interest.