Pubertal development and reproductive hormone levels of singleton ICSI offspring in adolescence: results of a prospective controlled study.

Abstract

Are there any differences in the pubertal development and reproductive hormone status during adolescence between singletons following ICSI therapy or spontaneous conception (SC)? Pubertal development and reproductive hormone levels are largely similar between ICSI and SC adolescents, except for a tendency towards lower inhibin B levels as well as significantly higher estradiol levels and a lower testosterone-to-estradiol-ratio in male adolescents. Previous data are scarce and partly inconclusive regarding pubertal development in female ICSI adolescents as well as demonstrating a tendency towards lower inhibin B serum levels in male ICSI offspring. Prospective controlled study including 274 singleton ICSI-conceived adolescents (141 girls, 133 boys) followed up for the third time, and 273 SC controls (142 girls, 131 boys) from seven German registration offices (Aachen, Eichstätt, Erfurt, Lübeck, Hamburg, Heidelberg and Schwerin). Pubertal development assessed by Tanner staging (breast, genital and pubic hair development), age at menarche and reproductive hormone levels were analyzed in ICSI and SC adolescents at the mean age of 16.5years. Differences were analyzed by multinomial regression (Tanner stages) or t test and linear regression for hormonal assessments. Both female and male ICSI and SC adolescents showed adequate pubertal maturation according to their age, and the mean age at menarche (at 12.7 versus 12.8years) was similar. Tanner stages as well did not display any relevant or significant differences between the groups. Reproductive hormone levels in female adolescents not using hormonal contraception were largely similar before and after adjustment for several factors such as preterm birth, Tanner stages, BMI or physical activity. In male ICSI adolescents, a tendency towards lower inhibin B (-14.8pg/ml, 95% CI: -34.2 to 4.6pg/ml), significantly higher estradiol (2.6ng/l, 95% CI: 0.0 to 5.2ng/l) and a significantly lower testosterone-to estradiol ratio (-0.047, 95% CI: -0.089 to -0.004) was found. The all-over response rate and the willingness to participate in the blood test and medical examination were very low in the control group. Participating control families may have greater health awareness, and selection bias cannot be ruled out. Hormonal data in the females were measured irrespective of the cycle day and restricted to those not using hormonal contraception. Some parameters from the questionnaire data such as usage of hormonal contraception might suffer from reporting bias. As this is an observational study, we can draw only limited causal conclusions from the findings. Differences in male reproductive hormones may indicate altered testicular function. However, at this time possible consequences for later reproductive success are unknown. DFG research grant KA 1643/4-1. The authors declare no conflict of interest.