Relationship between umbilical cord sex hormone binding globulin, sex steroids, and age at menarche: a prospective cohort study

The association between newborn cord blood steroids and ambient prenatal exposure to air pollution: findings from the ENVIRONAGE birth cohort

Knowledge of whether prenatal exposure to ambient air pollution disrupts steroidogenesis is currently lacking. We investigated the association between prenatal ambient air pollution and highly accurate measurements of cord blood steroid hormones from the androgenic pathway.This study included 397 newborns born between the years 2010 and 2015 from the ENVIRONAGE cohort in Belgium of whom six cord blood steroid levels were measured: 17α-hydroxypregnenolone, 17α-hydroxyprogesterone, dehydroepiandrosterone, pregnenolone, androstenedione, and testosterone. Maternal ambient exposure to PM2.5 (particles with aerodynamic diameter ≤ 2.5 μm), NO2, and black carbon (BC) were estimated daily during the entire pregnancy using a high-resolution spatiotemporal model. The associations between the cord blood steroids and the air pollutants were tested and estimated by first fitting linear regression models and followed by fitting weekly prenatal exposures to distributed lag models (DLM). These analyses accounted for possible confounders, coexposures, and an interaction effect between sex and the exposure. We examined mixture effects and critical exposure windows of PM2.5, NO2 and BC on cord blood steroids via the Bayesian kernel machine regression distributed lag model (BKMR-DLM).An interquartile range (IQR) increment of 7.96 µg/m3 in PM2.5 exposure during pregnancy trimester 3 was associated with an increase of 23.01% (99% confidence interval: 3.26-46.54%) in cord blood levels of 17α-hydroxypregnenolone, and an IQR increment of 0.58 µg/m³ in BC exposure during trimester 1 was associated with a decrease of 11.00% (99% CI: -19.86 to -0.012%) in cord blood levels of androstenedione. For these two models, the DLM statistics identified sensitive gestational time windows for cord blood steroids and ambient air pollution exposures, in particular for 17α-hydroxypregnenolone and PM2.5 exposure during trimester 3 (weeks 28-36) and for androsterone and BC exposure during early pregnancy (weeks 2-13) as well as during mid-pregnancy (weeks 18-26). We identified interaction effects between pollutants, which has been suggested especially for NO2.Our results suggest that prenatal exposure to ambient air pollutants during pregnancy interferes with steroid levels in cord blood. Further studies should investigate potential early-life action mechanisms and possible later-in-life adverse effects of hormonal disturbances due to air pollution exposure.

Mendelian randomization identifies age at menarche as an independent causal effect factor for gestational diabetes mellitus

AIMS

The relationship between age at menarche (AAM) and gestational diabetes mellitus (GDM) risk is still inconclusive. This Mendelian randomization (MR) analysis was used to assess systematically the causal relationship between AAM and GDM risk in human beings.

MATERIALS AND METHODS

Single-nucleotide polymorphisms associated with AAM, oestradiol levels, sex hormone-binding globulin (SHBG) levels and bioavailable testosterone (BioT) levels were screened via the genome-wide association study enrolling individuals of European descent. Summary-level data for GDM (123 579 individuals) were extracted from the UK Biobank. An inverse-variance-weighted method was used for the primary MR analysis. Sensitivity analyses were examined via MR-Egger regression, heterogeneity tests, pleiotropy tests and leave-one-out tests. The directionality that exposure causes the outcome was verified using the MR-Steiger test.

RESULTS

Genetically predicted early AAM was found to have a causal positive association with a higher risk of GDM (odds ratio = 0.798, 95% confidence interval = 0.649-0.980, p = .031). In the multivariable MR analysis adjusted for oestradiol, SHBG and BioT levels, the causal association between AAM and GDM risk remained (odds ratio = 0.651, 95% confidence interval = 0.481-0.881, p = .006). A 1-SD increase in SHBG or BioT levels was significantly associated with a 41.4% decrease or 20.8% increase in the overall GDM risk (p = 3.71E-05 and .040), respectively. However, after controlling for AAM, oestradiol levels and BioT levels by multivariable MR analysis, there was no direct causal effect of SHBG levels on GDM risk (p = .084). Similarly, after adjusting for AAM, oestradiol levels and SHBG levels by multivariable MR analysis, there was no direct causal effect of BioT levels on the risk of GDM (p = .533). In addition, no direct causal association was identified between oestradiol levels and GDM risk in univariable MR analysis or multivariable MR analysis.

CONCLUSION

Genetic variants predisposing individuals to early AAM were independently associated with higher GDM risk. Further research is required to understand the mechanisms underlying this putative causative association. In addition, AAM may be helpful in clinical practice to identify women at risk for GDM; pregnant women who are young for menarche may need to take precautions before GDM develops.

The levelling-off of the secular trend of age at menarche among Italian girls

OBJECTIVE

The main aim of this study was to verify whether the secular trend stopped in Italy by comparing the results of a 1990-2000 birth cohort versus a 1980-1990 birth cohort of Italian young women. The results were used to speculate about age at menarche as adaptive response to non-genetic factors.

METHODS

In 2016, a study was set on 413, 18-to-26 year-old women (1990-2000 birth cohort) attending two Italian Universities by web-based, self-reported questionnaires. Previously in 2000, a research including 3,783 high school female students (1980-1990 birth cohort) was led. The age at menarche distribution was performed by Kaplan-Meier analysis. The comparison between the findings of the two birth cohorts was performed by Wilcoxon sum-rank test. Mixed models analysis was applied to evaluate the effect of cohort and socio-economic status on age at menarche.

RESULTS

1990-2000 cohort's age at menarche median was 12.44y (95%CI 12.37; 12.59y). There was no significant difference with age at menarche of the previous cohort (p = 0.56). Consistently, the advance of age at menarche in comparison to the mothers' one was not significantly different between the two cohorts (-0.27y±0.10y vs -0.25y±0.03y, p = 0.33). The socio-economic level was not significantly associated with menarcheal age.

CONCLUSIONS

The findings of this study confirm that, like in other developed countries, the advance of age at menarche has stopped in Italy, consistently with the stop of the improvement of socio-economic conditions. Further studies are needed to explore the differential effect of each non-genetic factor to outline future scenarios of human sexual maturation.

TRIAL REGISTRATION

the Comitato Etico per la Sperimentazione Clinica (CESC) della Provincia di Padova of the Veneto Region (Italy), n°3993/U16/16.