Children conceived through frozen embryo transfers prepared with hormone replacement therapy tend to score about the same on early childhood intelligence tests as those conceived using natural cycle protocols. The hormone method was, however, linked to a higher rate of unusually large babies and more admissions to newborn care units. The research was published in Reproductive Biology and Endocrinology.
Assisted reproductive technology provides medical procedures to help people achieve pregnancy, and advancements in freezing techniques have made frozen embryo transfer increasingly common. In this procedure, an embryo created in a laboratory is frozen, thawed at a later date, and then placed into the mother’s uterus. Preparing the lining of the uterus to receive the embryo is a necessary step. The timing must align closely so that the uterus is ready when the embryo is introduced.
For women who have regular menstrual periods, doctors generally use one of two methods to prepare the uterus. The natural cycle method relies on the woman’s own hormonal cycle to naturally thicken the uterine lining. Medical staff monitor this cycle, sometimes use an injection to trigger ovulation, and then provide a supplemental progesterone-like drug to support early pregnancy.
In contrast, the hormone replacement therapy method lacks natural ovulation. Instead, patients take external hormones, specifically estrogen and progesterone, to artificially build up the lining and prepare it for the embryo.
While hormone replacement therapy is widely used, it relies on adding estrogen from medication, which alters the mother’s internal hormone levels. Prior studies provide evidence that being exposed to unusually high levels of estrogen during early pregnancy might be linked to changes in an offspring’s brain development.
The research, led by Xiaomeng Shi, Minyue Tang, and Yimin Zhu of Zhejiang University, aimed to compare the cognitive function of children conceived through these two different uterine preparation methods. By analyzing past intelligence assessments of children between three and six years old, an age range the authors note is too young for puberty to affect the results, the researchers hoped to provide better guidance for patients choosing an embryo transfer protocol.
The research team looked at medical records and follow-up assessments from a reproductive center in China. Out of 1,538 recruited families, 790 children born from single, non-twin pregnancies met the criteria and made the final sample. These children were born to mothers who had regular menstrual cycles and underwent frozen, early-stage embryo transfers between 2004 and 2017.
Blastocyst transfers, which involve embryos developed a few days longer in the laboratory, were excluded from the analysis. Out of the included children, 434 were conceived using the natural cycle method, and 356 were conceived using the hormone replacement therapy method.
At the follow-up visits, trained assessors evaluated the children’s cognitive function using a standardized intelligence test adapted for Chinese children. The test measured verbal intelligence through tasks like vocabulary and comprehension, and performance intelligence using mazes, block design, and a matching task. Together, these scores provided a full-scale intelligence quotient, or IQ. To reduce bias, the assessors did not know which embryo transfer method each child’s mother had used.
The researchers also gathered extensive medical data to account for variables that might influence a child’s intelligence. In their statistical model, they adjusted for the year of transfer, the parents’ ages and body mass index, a measure of weight relative to height. They also accounted for the fertilization method, how long the embryo was frozen, uterine lining thickness, the mother’s estrogen level, length of pregnancy, and newborn unit admissions.
The results indicated that the cognitive function of children born after hormone replacement therapy was similar to that of children born after the natural cycle method. The researchers found no statistically significant differences in verbal intelligence, performance intelligence, or full-scale IQ scores between the two groups. After adjustment, the hormone group’s full-scale IQ averaged about two points lower, a gap that fell just short of statistical significance. The results argue against a large difference but cannot rule out a small one.
Rates of what the researchers defined as intellectual disability, an IQ below 80, were similar regardless of how the mother’s uterus was prepared. This threshold is higher than the usual clinical cutoff of about 70. After adjustment, the hormone group scored slightly higher on the vocabulary subtest. The authors did not correct for running many comparisons and describe these secondary results as exploratory, so this single difference may reflect chance.
The scientists also broke the data down by the child’s sex and by the specific fertilization method used. They looked at sex because biological differences can sometimes lead to different sensitivities to the environment inside the womb. They looked at the fertilization method, separating standard in vitro fertilization from intracytoplasmic sperm injection, in which a single sperm is injected directly into an egg.
Earlier studies disagree on whether this injection procedure is tied to cognitive differences. In all of these smaller groups, the researchers still found no statistically significant differences in IQ scores between the two uterine preparation protocols.
The authors propose that estrogen levels in the hormone therapy group may fall below the levels tied to cognitive problems in earlier research, though the study did not test this directly. The typical level in the hormone group, measured in the mother’s blood shortly before the transfer, was about a third higher than in the natural cycle group. However, levels varied much more widely, and about a quarter of mothers in the hormone group had levels more than four times the typical natural cycle level.
Looking at the birth records, the study noted that babies in the hormone therapy group had a higher chance of being born unusually large for their gestational age, meaning heavier than 90 percent of babies born at the same point in pregnancy. Specifically, 22.2 percent of the babies in the hormone group were classified as large, compared to 15.2 percent in the natural cycle group. The babies in the hormone group also had a higher rate of admission to newborn care units, at 14.6 percent versus 9.0 percent for the natural cycle group. The major reasons for these newborn care unit admissions included jaundice, respiratory system diseases, and preterm birth.
As with all research, there are some caveats to consider. Because the study looked back at past medical records rather than randomly assigning patients to a protocol, unmeasured factors such as the parents’ lifestyle, family socioeconomic status, or the child’s early educational environment might influence the results. The timeline of treatments was also not evenly distributed, with hormone therapy becoming much more common in the later years of the study period. The analysis adjusted for transfer year, but changes in clinical and laboratory practice over that period could still affect the comparison.
There is also a chance of over-adjustment in the statistical model. By controlling for factors like estrogen levels and length of pregnancy, which are part of the pathway being studied or possible downstream effects, the model could potentially hide a real effect. The unadjusted results also showed no statistically significant differences, so the main conclusion does not depend entirely on these adjustments.
Additionally, the findings may not apply to blastocyst transfers, which are now common in modern clinics, or to women who do not ovulate regularly. Finally, the children were followed only up to age six. Future studies could track cognitive abilities into the teenage years to see if any subtle differences emerge over time.
The study, “Cognitive function in offspring born after frozen embryo transfer following hormone replacement therapy cycles versus natural cycles: a retrospective cohort study,” was authored by Xiaomeng Shi, Xiaojing Wu, Aqiong Xie, Shuangying Liu, Xiaoqian Zhou, Yingzhi Yang, Minyue Tang, and Yimin Zhu.