Question explored with the scientific record
Did COVID vaccines cause miscarriages
The evidence does not show a clear, consistent increase in miscarriage risk from COVID-19 vaccination, but the studies have important limits.
The largest study here, a 2025 Iranian cohort of 459 pregnancies, found 50 miscarriages (about 11 in 100), which is within the typical background rate [1]. A 2024 Spanish case-crossover study of 4,054 vaccinated pregnancies also found no overall association, though it noted a signal in the one-dose group that did not reach statistical significance [3]. A 2025 BMC study even reported a lower miscarriage rate in vaccinated women, but that study used only one vaccine brand (Sinopharm) and was retrospective [2].
The evidence is not clean. Most studies are observational, not randomized, and many compare vaccinated women to other vaccinated women rather than to a true unvaccinated control group. The VAERS-based safety reports [5] rely on a passive system that captures only a fraction of real events. The 2022 meta-analysis of 56,428 women found a relative risk of 1.23 for spontaneous abortion, but the confidence interval crossed 1.0, meaning the result was not statistically significant [4]. That is not proof of safety; it is an estimate too imprecise to rule out either benefit or harm.
| Study | Design | Key miscarriage finding |
|---|---|---|
| Iranian cohort 2025 [1] | Prospective, 459 pregnancies | 10.9% rate, within background |
| Spanish case-crossover 2024 [3] | 4,054 vaccinated pregnancies | No overall association; one-dose signal |
| BMC retrospective 2025 [2] | 610 women, Sinopharm only | Lower rate in vaccinated group |
| Meta-analysis 2022 [4] | 56,428 women, 3 studies | RR 1.23, not statistically significant |
My call: the best available evidence does not prove that COVID-19 vaccines cause miscarriages, but the studies are not designed to rule out a small or subgroup-specific risk. Confidence: moderate for the conclusion of no large excess risk; low for the conclusion of no risk at all, because the most informative study design (randomized, with a true placebo group and long follow-up) was never done.
Sources used 5
-
Evaluating miscarriage incidence after COVID-19 vaccination
A large Iranian cohort event monitoring study across seven cities evaluated miscarriage incidence after COVID-19 vaccination and found no increased risk across four vaccine brands (AZD1222, Sinopharm, Sputnik V, CoVIran Barekat), with 50 miscarriages among 459 pregnancies (10.9%…
DOI: 10.1038/s41598-025-06904-y -
Comparing the maternal and neonatal outcomes in vaccinated and unvaccinated pregnant women against COVID-19: a retrospective cohort study
This retrospective cohort study compares maternal and neonatal outcomes between vaccinated and unvaccinated pregnant women against COVID-19, revealing that vaccination significantly reduces the risk of miscarriage but increases the likelihood of neonatal jaundice.
DOI: 10.1186/s12884-025-07462-x -
Lack of association between COVID-19 vaccines and miscarriage onset using a case-crossover design
COVID-19 vaccination during pregnancy did not show an overall association with miscarriage onset in a case-crossover analysis of 607 miscarriages among 4,054 vaccinated pregnancies in Spain, though signals in the one-dose group merit further investigation.
DOI: 10.1038/s41598-024-57880-8 -
Efficacy and adverse effects of COVID-19 vaccine in pregnant women
Maternal mRNA COVID-19 vaccination appears generally safe in pregnancy, facilitates transplacental antibody transfer, and is linked to reduced COVID-19–related hospitalization in infants under 6 months, with mostly mild adverse effects.
DOI: 10.56294/saludcyt2022210 -
Safety of Booster Doses of Coronavirus Disease 2019 (COVID-19) Vaccine in Pregnancy in the Vaccine Adverse Event Reporting System
In VAERS, review of 323 adverse event reports in pregnant people after mRNA COVID-19 booster doses found no new or unexpected safety concerns, with pregnancy-specific and non-pregnancy-specific event patterns comparable to primary-series reports.
DOI: 10.1097/aog.0000000000004889