Researchers who read the complete genomes of 24,030 people from 7,851 parent and child families reported on August 7, 2026 in Nature Medicine that particular steps inside the fertility lab, and the age of both parents, leave a measurable mark on a baby’s DNA. Across those families the team catalogued 390,924 de novo single-nucleotide variants, meaning single-letter DNA changes present in the child but in neither parent.
Key takeaways
- The study sequenced 24,030 whole genomes and found 390,924 new single-letter mutations that the children did not inherit from either parent.
- Father’s age and mother’s age left different mutation patterns, and the mother’s mutation count sped up at older ages.
- Different fertility procedures had different fingerprints, and the effects held independent of parental age.
What the study found
Paternal and maternal aging produced distinct mutational patterns, according to the study, with maternal mutation buildup accelerating at advanced ages. The authors also report that the extra paternal mutations partly explained why older parents tended to have shorter pregnancies. In other words, the mutation count was not just sitting alongside the age effect, it appeared to be part of the path from age to a shorter gestation.
The assisted reproduction findings are the part that will get repeated. The authors describe the effects as age independent and procedure specific. Intracytoplasmic sperm injection, known as ICSI, is the technique where a single sperm is injected directly into an egg, and it was associated with more paternal mutations. Ovarian stimulation, the hormone phase that produces multiple eggs, was associated with more maternal mutations. And ICSI-linked paternal mutations partly explained the association between ICSI and a shorter pregnancy.
Handling the embryo in the lab was associated with more early post-zygotic mosaic mutations, changes that arise after fertilization so they show up in only some of the child’s cells. One class in particular, C to A substitutions, was linked to delayed neurocognitive development at 1 year of age.
Dr. Kumar’s take
Read the headline version of this and you get “IVF damages babies’ DNA.” That is not what the paper shows, and the gap matters.
What was measured is mutation count. What was not measured is a rise in diagnosed disease. Those are different things. Most de novo mutations land in stretches of DNA where nothing happens. A higher average count across thousands of children can be real and still leave any individual child’s risk close to where it started.
The one-year neurodevelopmental signal deserves the same discipline. Developmental scoring at 12 months is noisy. What the paper reports is an association between one class of mutation and delayed neurocognitive development at 1 year, assessed once, early. It is worth following. It is not worth a decision.
The useful finding here is the specificity. This is not IVF as a category. It is ICSI, and ovarian stimulation, and physical embryo handling, three separate steps with three separate signatures. That is what makes the biology credible, and it is also what makes it actionable, because those steps are chosen. ICSI in particular gets used well beyond the male-factor infertility it was designed for.
And then there is the part nobody brings up in the clinic. Parental age showed up in this data too, with its own pattern and an apparent route to shorter pregnancy. Age is the variable most couples control least and hear about least, at least on the father’s side. If mutation burden is the worry, age belongs in that conversation before the lab protocol does.
This is observational work. It cannot prove that any procedure caused the mutations it is associated with, and couples who need ICSI differ from couples who do not in ways that are hard to fully adjust away.
What it means for you
If you are in fertility treatment or considering it, this is not a reason to walk away from IVF. It is a reason to ask which specific steps your clinic plans to use and why. Whether you need ICSI, the single sperm injection step, is a real question worth asking, especially if there is no male-factor reason for it.
If you are deciding when to start, treat parental age as a shared factor rather than a maternal one. The same study found paternal patterns too.
For everything else that shapes a pregnancy, the more established levers still apply. Everyday chemicals in pregnancy have been linked to early birth, and ultra-processed food may lower fertility and slow early embryo growth. Those are worth your attention now. A mutation count in a research database is worth watching, not fearing.