Frozen embryo transfer linked to twice the childhood cancer rate in 1.1 million Danish children

Fertility treatment has quietly changed shape over the past two decades. Most embryos made in the laboratory are now frozen and thawed before they are transferred, and a technique that injects a single sperm into each egg has become routine even for couples without male infertility. A long-running Danish study suggests those two procedures are linked to a higher rate of childhood cancer, though cancer in children remains rare.

A research team led by Katrine Mose, with Marie Hargreave as senior author, followed every child born alive in Denmark from 1996 to 2012, more than a million in all, for a median of 20 years. Children conceived by frozen embryo transfer developed cancer at about twice the rate of children conceived without fertility help, and those conceived by intracytoplasmic sperm injection, known as ICSI, at a rate about 50% higher. Children from conventional IVF, the original method in which egg and sperm simply meet in a lab dish, had the same rate as everyone else. Writing in the journal JAMA on October 21, the authors stress that absolute risks were low.

What the three IVF techniques are

All three methods begin the same way, with eggs collected from a woman and fertilized in a laboratory. In conventional IVF, eggs and sperm are placed together and fertilization happens on its own. In ICSI, a single sperm is injected directly into each egg. It was developed for male infertility, but its use has spread far beyond that. Among fresh IVF cycles in the United States, a national analysis found, the share using ICSI rose from 36.4% in 1996 to 76.2% in 2012. Most of the growth came in couples without male factor infertility, among whom ICSI use rose from 15.4% to 66.9% of cycles over those years, and in them it brought no better pregnancy results than conventional IVF.

In frozen embryo transfer, an embryo is frozen after fertilization and thawed for transfer in a later cycle, sometimes long after it was made. Freezing has spread even faster, helped by better survival of frozen embryos. In the United States, frozen-thawed embryos accounted for 78.8% of all embryo transfers using a woman’s own eggs by 2019, a figure cited by a Nordic research group, and a similar pattern is observed in Australia, New Zealand and Europe. Some clinics now freeze every embryo by default, a policy the Nordic group says is spreading without careful consideration of benefits and harms.

What is new since the 2019 Danish report

The same Danish group raised the frozen-embryo question in JAMA in 2019, using the same birth years. That analysis followed the children for an average of 11.3 years and found that frozen embryo transfer was associated with a raised risk of childhood cancer: 44.4 cases per 100,000 children a year, against 17.5 for children born to fertile women. That signal rested on only 14 cancer cases after frozen transfer. The excess was mainly due to an increased risk of leukemia and of tumors in nerve tissue outside the brain, and the other treatments, including ICSI, showed no link that could be told apart from chance.

The new paper confirms that signal and extends the work in four ways. Follow-up now runs through 2023, a median of 20 years, so cancers of the teenage years and early adulthood are counted (by 2023 the oldest children were in their late 20s, though the paper reports its results as childhood cancer), and the total reached 3,736 cancers against 2,217 in the earlier analysis. ICSI now shows a link of its own. The results are broken down by type of cancer. And the comparison is repeated against two other groups of children, those whose mothers had a history of infertility and those conceived by conventional IVF, to test whether infertility itself explains the pattern. The frozen-embryo estimate came down a little from the 2019 figure, but it held.

How much more common cancer was after frozen embryo transfer and ICSI

Of the 1,086,136 children in the study, 37,228 (3.4%) were conceived using assisted reproduction. Among children conceived without it, cancer appeared at a rate of 18.3 cases per 100,000 children per year. Across all assisted reproduction the rate was 23.6, almost 30% higher.

That rise came from two of the three techniques. After frozen embryo transfer the rate was 41.1 per 100,000 a year, roughly double. After ICSI it was 27.9, about 50% higher. After conventional IVF it was 18.5, effectively the same as in children conceived without help.

In absolute terms the extra risk is small. The gap after frozen transfer works out to about 23 extra cases per 100,000 children per year, so for almost every family childhood cancer remains an unlikely event either way.

The authors also looked at which cancers drove the pattern. Frozen embryo transfer was associated with leukemia, lymphomas and soft tissue sarcomas, which are cancers of muscle and connective tissue. ICSI was linked to lymphomas, soft tissue sarcomas and tumors of the brain and spinal cord. For conventional IVF, no associations were observed with any cancer type. Each of these cancers is uncommon, so the estimates for single types come with very wide margins of error.

The most telling check concerns infertility itself. Couples who need fertility treatment may differ from others in ways that could affect a child’s cancer risk. When the researchers compared ICSI and frozen-transfer children with children of women who had a history of infertility, or with children conceived by conventional IVF, the results were similar. In the authors’ words, this suggests “procedure-related, rather than infertility-related, factors warrant further investigation.”

How the Danish results fit with other studies

The wider evidence is mixed. A study of children born in four Nordic countries, 7.9 million in all, found no overall rise in cancer after assisted reproduction. Yet children born after frozen embryo transfer had a higher risk of childhood cancer than those born after fresh transfer or natural conception: 30.1 cases per 100,000 a year, against 18.8 after fresh transfer and 16.7 after natural conception. Leukemia again stood out. In the Nordic data a higher risk was observed for leukemia after frozen transfer than after fresh transfer, at 14.4 against 6.2 cases per 100,000 children a year, and the gap held after the researchers allowed for birth weight and birth defects. That study includes Danish children from overlapping years and Danish researchers, so it is not a fully separate test. Very high birth weight is more common after frozen transfer and is itself linked to childhood cancer, but adjusting for it changed the Nordic results only marginally.

Other countries have not seen the same pattern. In Britain, 106,013 children born after assisted conception had almost exactly the number of cancers expected, with no overall rise, though two rare tumors, of the liver and of muscle, were more common. In the United States, a study of 275,686 children conceived via IVF found a small overall rise in early childhood and more liver tumors, but no link with specific treatment methods. Both followed children for a shorter time than the Danish study.

A 2025 meta-analysis of 18 studies, restricted to children whose parents had infertility, found no significant link between assisted reproduction and childhood cancer, including for ICSI and frozen transfer, and rated the quality of that evidence as low to moderate. The Danish data, with 20 years of follow-up and the infertility comparison built in, now sit on one side of that picture.

Why a procedure might matter is unknown. The Nordic authors note that high birth weight has been tied to childhood cancer, and that epigenetic changes, chemical tags on DNA that control which genes are switched on, have been proposed as a possible explanation. They add that each childhood cancer has its own risk profile and that many are thought to begin in the womb.

What the Danish IVF study cannot show

This is an observational study, so it shows a link and cannot prove that frozen embryo transfer or ICSI causes cancer. Couples who receive these treatments differ in age, health and the reasons they need help, and comparing against infertile and IVF families narrows that gap without closing it.

The numbers within each technique and cancer type are small, which is why the estimates for single cancers range so widely. The children were born between 1996 and 2012, a period when freezing methods and laboratory practice differed from today’s, so the results may not describe current techniques.

What the Danish IVF results mean for families

For couples weighing fertility treatment, the study sharpens a question more than it settles one. Frozen transfer and ICSI are now the norm in many clinics, and the Danish authors call for further investigation of procedure-related factors. Which treatment suits a particular couple is decided with their fertility specialist, who can weigh this evidence against each technique’s benefits.

For children already born after these procedures, the figures describe rates across a whole population, about 41 cancers per 100,000 children a year after frozen embryo transfer against about 18 in children conceived without help, and most of those children will never have cancer.