TL;DR: A new paper in Nature Communications looked at THC around eggs during IVF and saw a small dip in the share of embryos that look chromosomally “normal” in the lab. A local TV piece turned that into “weed causes infertility, miscarriages, and genetic defects.”
That’s not what the study showed. It didn’t track pregnancies or babies and used leftover immature eggs in dishes. Meanwhile, big human studies find no slower time-to-pregnancy and an IVF cohort found no worse IVF results. If you’re trying to conceive or are pregnant, mainstream guidance says to pause use, per the CDC and ACOG. The scary headlines go further than the data.
What the new paper actually did
- Clinic records at one IVF center: They checked the fluid around eggs for THC leftovers. A small group tested positive. Those patients had a slightly lower rate of embryos that look chromosomally normal in the lab (60% vs 67%). That’s a lab checkpoint, not a pregnancy.
- Dish experiments on leftover immature eggs: They put unused, not-fully-mature eggs in dishes with THC and watched. Some lab signals moved in a worrying direction (more “wonky” chromosome-sorting gear at a higher dose), but several results—like more eggs with the wrong number of chromosomes—were too small and noisy to be sure.
Bottom line: This shows a possible lab signal, not proof that cannabis users struggle to get pregnant or have more miscarriages.
Why the headlines overshot
The TV piece said “infertility, miscarriages, genetic defects.” The paper didn’t measure any of those. It measured proxies in a lab (how embryos look under a microscope). Proxies can be useful, but they don’t equal real-world outcomes.
The fine print (from the paper itself)
- Tiny exposed group: Only a small slice of IVF patients tested THC-positive.
- What moved: The % of embryos that looked “chromosomally normal” dipped (60% vs 67%). Fertilization and blastocyst development didn’t clearly get worse.
- What didn’t stick statistically: More eggs with chromosome mistakes? The numbers were not solid enough to rule out chance. The “weird spindle shape” finding hit significance only at the higher, lab-style dose with very small groups.
- What they didn’t measure: No dose, timing, or type of cannabis. No pregnancy, miscarriage, or live-birth results.
Who funded it, and why that matters
The authors note University of Toronto fellowship support and funding from the CReATe Fertility Centre (via reinvested clinical earnings). That isn’t “Big Pharma,” but it is a fertility clinic—so there can be an incentive to counsel toward caution and additional testing. The authors report no competing interests. Either way, replication by independent teams is how we find the truth.
What the rest of the evidence says (humans first)
- Trying naturally: A large pre-pregnancy study found no slower time-to-pregnancy for cannabis users.
- In IVF clinics: A real-world cohort found no worse IVF outcomes (eggs retrieved, fertilization, embryo development, pregnancies).
- Pregnancy outcomes: Some studies raise concerns (e.g., low birthweight), but many rely on self-report and are tangled up with tobacco, alcohol, stress, and other confounders. Newer work on placental-related outcomes flags caution without proving causation. This is why mainstream orgs say avoid cannabis during pregnancy.
What this paper does not prove
- It does not prove that cannabis causes infertility.
- It does not show higher miscarriage rates.
- It does not show birth defects in babies.
- It does not report live births or take-home baby rates.
It raises a biological question that bigger, better studies need to answer.
How to read studies like this (no PhD required)
- People ≠ petri dishes: Lab results can hint at risk, but real life is messier.
- Proxies vs outcomes: Lab grades of embryos aren’t the same as pregnancies and babies.
- Small numbers swing wildly: Tiny groups make “significant” blips easy to oversell.
- Confounders are everywhere: THC-positive patients may differ in sleep, stress, tobacco, alcohol, BMI, and timing—matching helps, but can’t fix it all.
- Follow the incentives: Funding doesn’t make research wrong, but it can shape what gets studied and how it’s framed.
What would actually settle it?
- Big, multi-center, preregistered human studies that measure how much, how often, and how people use.
- Track what matters: pregnancy, miscarriage, live birth, newborn health—not just embryo grades.
- Use the lab stuff (genes, spindles, chromosome checks) as secondary clues, not headlines.
- Independent teams, not only clinic-funded groups.
Practical takeaways
- If you’re in an IVF cycle or pregnant: Best practice is to pause cannabis until after pregnancy and breastfeeding, per guidance from the CDC and ACOG.
- If you’re trying to conceive: Current human data show no slower time-to-pregnancy and no worse IVF outcomes. Still, being cautious around key windows (stimulation, retrieval, transfer) is reasonable.
- If you’re just headline-doomscrolling: One study ≠ settled science. Wait for replication.
Quick glossary
- Euploid embryo = embryo with the usual number of chromosomes (a lab “looks normal” grade).
- Aneuploidy = wrong number of chromosomes (many won’t implant or will miscarry).
- Spindle = the egg’s internal “sorting machine” that splits chromosomes during maturation.
Editor’s note: If you’re pregnant, breastfeeding, or in an IVF cycle, major medical orgs advise avoiding cannabis and talking with your clinician.
Medical info disclaimer: This article is journalism, not medical advice.
Photo by Noah Buscher on Unsplash
