A first-trimester blood protein flags preeclampsia risk months early
Screening 2,904 proteins in early pregnancy, researchers found low maternal isthmin-2 was the strongest signal for later preeclampsia or fetal growth restriction, validated in two cohorts.
Preeclampsia and fetal growth restriction are among the largest causes of maternal and newborn harm worldwide, and both trace back to the same early failure: the placenta's specialised cells do not invade the wall of the uterus as deeply as they should. A study in Nature Medicine, published 1 September, reports a blood protein measured at around 12 weeks of pregnancy that may flag that failure long before either condition becomes apparent [s1].
What the researchers did
The team ran proteomics on maternal serum drawn at roughly 12 weeks' gestational age in a prospective pregnancy cohort, the Pregnancy Outcome Prediction Study, screening across 2,904 proteins to see which best foreshadowed later preeclampsia or fetal growth restriction [s1]. The strongest signal was low maternal serum isthmin-2 (ISM2): among nearly three thousand candidates, it stood out as the single best first-trimester marker for those outcomes [s1].
Crucially, they did not stop at one cohort. The association was validated in two independent groups — a second Pregnancy Outcome Prediction Study and the Improving Maternal Pregnancy And Child ouTcomes study [s1]. Replication in separate populations is what separates a durable biomarker from a statistical fluke, and it is the reason this result is worth attention.
The biology behind the number
The study goes further than a correlation, which is unusual and strengthens the case. ISM2 protein and its messenger RNA are produced almost exclusively in the placenta, and within the placenta the mRNA is highly enriched in extravillous trophoblast — the very cells that invade the uterus [s1]. In laboratory experiments, knocking down ISM2 in cultured human trophoblast stem cells profoundly inhibited their invasion, while forcing ISM2 expression in a cell line that normally lacks it promoted migration [s1]. That pairing — an epidemiological signal plus a plausible mechanism moving in the same direction — is why the authors suggest ISM2 may be causally involved in the early pathophysiology, not merely a bystander marker [s1].
An accompanying commentary in the same journal frames the finding as a possible window onto the early origins of both conditions [s1][s2].
What this is not, yet
It is not a test a pregnant person can ask for. This is a discovery-and-validation study, not a trial of a screening tool: the authors describe ISM2 and its pathways as potential targets for prediction and prevention, which is a statement about promise, not proven clinical utility [s1]. Several steps stand between here and a clinic. A predictive marker has to be turned into an assay with defined thresholds; its performance — how many true cases it catches and how many false alarms it raises — has to be measured prospectively; and, most important, a positive result only helps if there is something to do about it.
On that last point there is at least a foothold. Low-dose aspirin started in early pregnancy already reduces preeclampsia risk in high-risk women, so a first-trimester marker that identifies risk earlier and more accurately could, in principle, direct existing prevention better — but whether ISM2-guided screening actually improves outcomes is an untested hypothesis this study does not address [s1].
It is worth being clear about what "strongest signal" means here, and what it does not. Out of 2,904 proteins measured, ISM2 rose to the top as the single best first-trimester discriminator for these outcomes, which is a stringent bar to clear — but a marker can be the best of a large field and still not be accurate enough, on its own, to guide care [s1]. The paper does not report the kind of sensitivity-and-specificity performance that a screening test lives or dies by, because that was not its purpose; its purpose was discovery, replication and mechanism [s1]. The replication in two independent cohorts and the trophoblast experiments are what lift it above the many candidate biomarkers that surface in one dataset and evaporate in the next, and they are the reason the accompanying commentary treats ISM2 as a genuine lead into the early origins of the disease rather than a curiosity [s1][s2].
What to watch
The near-term questions are practical: can ISM2 be measured reliably and cheaply, does it add predictive value on top of the markers and clinical factors already used in first-trimester screening, and does acting on it change what happens to mothers and babies. The causal biology also invites a longer-term question — whether the ISM2 pathway itself could be a target for preventing, not just predicting, failed placentation. None of that is settled by one discovery study, however well replicated.
This article is informational and does not constitute medical advice.
Sources
- [s1] Isthmin-2 is a first-trimester predictor of preeclampsia and fetal growth restriction. Nature Medicine, 1 September 2026. https://doi.org/10.1038/s41591-026-04573-6
- [s2] Isthmin 2 could determine the early origins of preeclampsia and fetal growth restriction. Nature Medicine, 2 September 2026. https://doi.org/10.1038/s41591-026-04604-2
Sources
- Isthmin-2 is a first-trimester predictor of preeclampsia and fetal growth restriction — Nature Medicine , September 1, 2026
- Isthmin 2 could determine the early origins of preeclampsia and fetal growth restriction — Nature Medicine , September 2, 2026
More on
Women with a pre-eclampsia history had lower breast cancer risk, Finnish data show
A Finnish registry of over 123,000 women found pre-eclampsia linked to an 11% lower breast cancer risk decades later. Hormone therapy raised risk by a similar amount in both groups.
Doubling the aspirin dose did not prevent more preeclampsia in a 365-woman trial
The 162 mg group had 14.1% preterm or severe preeclampsia against 17.1% on 81 mg — a difference the trial could not distinguish from chance. Eight placental abruptions occurred on the higher dose and none on the lower.
Long COVID before or during pregnancy was not linked to preterm birth in a US cohort
Among 603 RECOVER participants, 118 were classified as likely having Long COVID. The primary analysis found no association with preterm delivery, and one secondary signal that only appeared in a sensitivity analysis.
No clear link between early-pregnancy GLP-1 use and hypertensive disorders
Pooling under 11,000 pregnancies, a meta-analysis found no statistically significant association either way, with the three underlying studies disagreeing on which direction the risk even points.