More than 13 million children have been born worldwide through assisted reproduction. Could the first days of life in a laboratory leave a trace in their DNA?
Scientists in China looked for one possible sign and found a difference in the genome's so-called 'jumping genes'.
Previous research has suggested that some health conditions may be slightly more common among children conceived through in vitro fertilization (IVF).
Finding out why is difficult. Any effects of treatment must be separated from factors such as their parents' health and what happens during pregnancy.
Researchers at Sun Yat-sen University turned their attention to pieces of DNA known as jumping genes.
Despite the name, they are not ordinary genes. Some can make copies of themselves and insert those copies into new parts of our DNA. The type examined in this study is called LINE-1.
An earlier study in mice had found that experiences early in life could affect jumping genes in the brain. What happens in mice, though, cannot tell us what happens to a human embryo during IVF.
To look for a difference, the team examined blood samples from 33 children conceived through IVF and 42 conceived without fertility treatment. Most samples were collected from umbilical cords at birth. Some came from children aged two to five.
The children conceived through IVF had a slightly higher amount of LINE-1 DNA overall.
The team also checked whether the result depended on the type of fertility treatment. Some children had been conceived through conventional IVF, while others were conceived through a procedure in which a single sperm is injected into an egg.
The researchers found no clear difference in LINE-1 amounts between those two groups.
The researchers had another way to explore it. In three families, they could compare a child conceived through IVF with a brother or sister conceived without fertility treatment.
The IVF-conceived child had more LINE-1 DNA in all three families. But three families are too few to establish that the difference is due to IVF.
When they separated the children by sex, the difference was clearer among girls. Boys showed a similar trend, but the result was less certain. The researchers say a larger group is needed to find out whether that distinction holds up.
When the team looked more closely at the children's DNA, they found that at some spots, LINE-1 sequences were detected more or less often in one group than the other. Some of those spots lay within a wide stretch of DNA around genes previously linked to the heart, metabolism, the brain, and cancer.
Could that help explain the health differences reported in earlier research? Perhaps, but this study cannot answer that question yet.
The team did not measure whether those genes behaved differently, and did not follow the children to see whether any became ill. For now, the connection gives scientists somewhere specific to look next.
The researchers also cannot say exactly when the difference arose. They read the children's DNA from blood samples; they did not watch jumping genes move in embryos.
Without DNA from both parents, they cannot tell whether particular changes appeared in the children or were passed down through the family. Other experiences during pregnancy may also have played a part.
A separate, much larger study published in Nature Medicine compared the DNA of 8,328 babies with that of their parents. It found links between particular fertility treatment steps and new DNA changes, some arising early in development.
That study examined a different kind of change, so it does not confirm the new finding, but it's yet more evidence that assisted reproduction may leave a mark.
In terms of the Chinese research, the team has uncovered a pattern that may help scientists ask more precise questions about IVF and early development.
The next step is to see whether it appears again in larger groups of families, where it comes from, and whether it makes any difference to children's health.
The research has been published in American Journal of Obstetrics & Gynecology.
This article was fact-checked by Peter Dockrill and edited by Peter Dockrill. While we pride ourselves on our process, we are only human. If you spot a mistake, please let us know.