Hi, you suggested freezing embryos instead of eggs. However, a recent study found that frozen embryo babies are 2.5 times more likely to develop cancer. At the same time, a review article on IVF comes to the following conclusion (likely due to epigenetic factors):
IVF is associated with an increased health risk for the offspring in the form of malformations, functional dis- orders, and a poorer peripartum outcome. This is possibly caused by parental factors, but also by factors related to the IVF technology used. [...] Consequently, IVF should only be performed if a pregnancy cannot be achieved in another way.
This makes me want to consider two points:
Assisted reproductive technology carries risks and couples that go through IVF should in any case do polygenic testing to offset these
Is the risk/benefit ratio worth it for couples that can conceive naturally, especially for something like merely 5 IQ points?
Thank you for leaving such a thought-provoking comment. I’ve spent a couple hours reading through the study you posted tonight as well as others linked to by the authors.
I don’t see the claim about a 2.5x increased risk of cancer anywhere though. From the findings section:
Children born after FET had a higher risk of cancer (48 cases; IR 30.1/100,000 person-years) compared to both fresh embryo transfer (IR 18.8/100,000 person-years), aHR 1.59, 95% CI 1.15 to 2.20, p = 0.005, and spontaneous conception, aHR 1.65, 95% CI 1.24 to 2.19, p = 0.001.
So the risk of cancer was 8% higher in those born after ART, and 59% higher for frozen embryos vs fresh embryos.
I think the generally higher disease prevalence among IVF couples probably explains the 8% increase for ART in general, though the 59% increase they see for frozen embryo transfer is surprising.
Looking more into the study it looks like about a quarter of the effect is driven by the higher rates of twin births in IVF, which are much less common nowadays.
This study uses data that is also quite old; they include cycles going all the way back to 1984 or 1994 for some countries. The rate of embryo freezing at that time were quite low, as evidence by the huge difference between hazard ratios for all ART and frozen embryo transfer. If frozen embryos made up a higher proportion of the births you would see a smaller difference between all ART relative to spontaneous conception and frozen embryo transfer relative to spontaneous conception.
Here’s another study that found higher risk of neoplasms among embryos that were transferred fresh. Granted, this was a smaller study, so I’d lean towards believing your study.
There’s also a graph in the study which seems to show the relative cancer risk for frozen embryos declining over time:
Though this could just reflect fewer twin births. And the confidence interals are such that it’s hard to be certain the effect is real.
Another possible confounder here is maternal age. The average age of mothers in ART were older than those in the spontaneous conception group by about 4 years. You can see in this study that maternal age is significantly associated with higher childhood cancer risk. So this could explain another 5-10%, particularly if the age of mothers in the frozen embryo group was higher (I didn’t see a comparison in the study but I could have missed it).
Still, there remains a chance that freezing embryos does increase childhood cancer risk. The absolute risk increase is still quite small: about 0.2% up to age 18. Though I suppose the increase in risk for adults could be higher?
Overall I’m just not sure what to think. This is one study and my experience with these “association studies” is they never control for confounders. Like what are the indications for embryo freezing and could those potentially account for this difference? Was childhood cancer risk the only thing the authors tested for, or are there other associations they didn’t report? I guess I’ll update my priors in favor of frozen embryo transfer increasing childhood cancer risk a bit?
Assisted reproductive technology carries risks and couples that go through IVF should in any case do polygenic testing to offset these
Yes, this one seems like a no-brainer to me, regardless of whether or not the effects we see are caused by IVF or merely associated with it by virtue of the parents that need it having higher polygenic risk scores for disease. The fact that PGT-P isn’t already universal in IVF is kind of a tragedy. You can avoid like $200k in future medical expenses for like $3-5k. Not to mention the improvement in quality of life that comes with lower risk of obesity, clinical depression, type 1 diabetes and other early onset conditions.
Is the risk/benefit ratio worth it for couples that can conceive naturally, especially for something like merely 5 IQ points?
If the effect is indeed real, I would happily take a .2% increase in childhood cancer risk for +5 IQ points. I don’t think that level of increased risk is really much of an issue for most parents. The much bigger issue is the cost of doing IVF. At $30k minimum, this is an expensive procedure.
I’m relatively confident there are at least a few tens of thousands of parents in the US who would do IVF for the purposes of these benefits if they knew this was possible. But for the time being with the benefits being relatively minor, the vast majority of parents will of course opt for natural birth.
With a better predictor capable of +5-13 IQ points I think that number will expand significantly. And if we can lower the cost of IVF I expect it to expand much farther. But we have to start somewhere.
Hi, you suggested freezing embryos instead of eggs. However, a recent study found that frozen embryo babies are 2.5 times more likely to develop cancer. At the same time, a review article on IVF comes to the following conclusion (likely due to epigenetic factors):
This makes me want to consider two points:
Assisted reproductive technology carries risks and couples that go through IVF should in any case do polygenic testing to offset these
Is the risk/benefit ratio worth it for couples that can conceive naturally, especially for something like merely 5 IQ points?
Thank you for leaving such a thought-provoking comment. I’ve spent a couple hours reading through the study you posted tonight as well as others linked to by the authors.
I don’t see the claim about a 2.5x increased risk of cancer anywhere though. From the findings section:
So the risk of cancer was 8% higher in those born after ART, and 59% higher for frozen embryos vs fresh embryos.
I think the generally higher disease prevalence among IVF couples probably explains the 8% increase for ART in general, though the 59% increase they see for frozen embryo transfer is surprising.
Looking more into the study it looks like about a quarter of the effect is driven by the higher rates of twin births in IVF, which are much less common nowadays.
This study uses data that is also quite old; they include cycles going all the way back to 1984 or 1994 for some countries. The rate of embryo freezing at that time were quite low, as evidence by the huge difference between hazard ratios for all ART and frozen embryo transfer. If frozen embryos made up a higher proportion of the births you would see a smaller difference between all ART relative to spontaneous conception and frozen embryo transfer relative to spontaneous conception.
Here’s another study that found higher risk of neoplasms among embryos that were transferred fresh. Granted, this was a smaller study, so I’d lean towards believing your study.
There’s also a graph in the study which seems to show the relative cancer risk for frozen embryos declining over time:
Though this could just reflect fewer twin births. And the confidence interals are such that it’s hard to be certain the effect is real.
Another possible confounder here is maternal age. The average age of mothers in ART were older than those in the spontaneous conception group by about 4 years. You can see in this study that maternal age is significantly associated with higher childhood cancer risk. So this could explain another 5-10%, particularly if the age of mothers in the frozen embryo group was higher (I didn’t see a comparison in the study but I could have missed it).
Still, there remains a chance that freezing embryos does increase childhood cancer risk. The absolute risk increase is still quite small: about 0.2% up to age 18. Though I suppose the increase in risk for adults could be higher?
Overall I’m just not sure what to think. This is one study and my experience with these “association studies” is they never control for confounders. Like what are the indications for embryo freezing and could those potentially account for this difference? Was childhood cancer risk the only thing the authors tested for, or are there other associations they didn’t report? I guess I’ll update my priors in favor of frozen embryo transfer increasing childhood cancer risk a bit?
Yes, this one seems like a no-brainer to me, regardless of whether or not the effects we see are caused by IVF or merely associated with it by virtue of the parents that need it having higher polygenic risk scores for disease. The fact that PGT-P isn’t already universal in IVF is kind of a tragedy. You can avoid like $200k in future medical expenses for like $3-5k. Not to mention the improvement in quality of life that comes with lower risk of obesity, clinical depression, type 1 diabetes and other early onset conditions.
If the effect is indeed real, I would happily take a .2% increase in childhood cancer risk for +5 IQ points. I don’t think that level of increased risk is really much of an issue for most parents. The much bigger issue is the cost of doing IVF. At $30k minimum, this is an expensive procedure.
I’m relatively confident there are at least a few tens of thousands of parents in the US who would do IVF for the purposes of these benefits if they knew this was possible. But for the time being with the benefits being relatively minor, the vast majority of parents will of course opt for natural birth.
With a better predictor capable of +5-13 IQ points I think that number will expand significantly. And if we can lower the cost of IVF I expect it to expand much farther. But we have to start somewhere.