An evidence library should include studies that fail to confirm a proposed biomarker, not only those with favorable results. This report found no statistically significant separation by serum kisspeptin. Primary study
What the researchers compared
Ninety women in a narrow early-pregnancy window were divided into healthy-pregnancy, threatened-miscarriage, and spontaneous-miscarriage groups of 30 each. Median kisspeptin concentrations did not differ significantly across the groups. The investigated blood-count measures also did not provide significant group separation in the reported comparisons. Primary report
Our appraisal: null is neither proof of equivalence nor a reason to omit the study
A nonsignificant result means the analysis did not establish a difference under its assumptions. It does not prove identical concentrations, and it does not establish that every clinically relevant effect has been excluded. Effect estimates and uncertainty would help determine what differences remain compatible with the data.
The group definitions matter. Threatened miscarriage and confirmed pregnancy loss are not interchangeable outcomes. A comparison of currently defined groups is also different from prospectively predicting a later event in people whose outcome is not yet known.
The narrow gestational window may improve timing comparability while limiting applicability outside that window. Differences from other reports could arise from timing, specimen type, assay behavior, population selection, or outcome definitions. Those possibilities should be investigated rather than resolved by selecting whichever result looks more persuasive.
Equal-sized groups are convenient for comparison but do not reproduce routine-care prevalence. Predictive values cannot simply be inferred from that balance. Nor does a lack of association in one small study establish a universally useless marker.
This report contributes a necessary counterpoint to positive biomarker studies. It does not support a clinical kisspeptin test in its studied setting and does not test peptide administration. A balanced synthesis should retain the null finding, evaluate methodological differences, and avoid treating association research as evidence that changing hormone levels changes pregnancy outcomes.
Frequently asked questions
How many participants were included?
Were groups equal in size?
Yes, 30 each.
Was a significant kisspeptin difference found?
No.
Does nonsignificance prove equivalence?
No.
Is threatened miscarriage identical to confirmed loss?
No.
Do equal study groups imply equal real-world prevalence?
No.
Why retain this report in the library?
It prevents a selectively positive account of an inconsistent biomarker literature.