A blood marker could make a difficult-to-sample physiological process easier to study. This report examined that possibility for INSL3 and intratesticular testosterone. Primary report
Study snapshot
The study involved 37 healthy men in a controlled experiment using acute gonadotropin suppression and hormonal interventions. After ten days, serum INSL3 correlated strongly with intratesticular testosterone. Other candidate serum markers did not show the same relationship in this setting. The authors proposed INSL3 as a potentially useful marker in some clinical contexts. Primary report
Our appraisal: correlation is the beginning of validation
A strong association does not show how accurately an individual patient's local hormone concentration can be predicted. Prediction error, calibration, and performance in a separate sample would be needed before treating the blood test as a replacement for the underlying measurement.
The intervention may also move both measurements together. That can produce a strong across-group association even when the relationship within each treatment group is less informative. A careful analysis should distinguish those possibilities rather than relying only on one pooled correlation.
The experimental context is another boundary. Acute suppression in healthy volunteers is not equivalent to chronic endocrine disease, and a marker's behavior may depend on that context. The useful next question is where the relationship holds, not whether the initial association sounds biologically plausible.
For evidence mapping, verify whether other publications report the same participants. A biomarker analysis can add a distinct question without representing another independent trial. Full review should inspect paired-sample completeness, assay performance, and prespecified analysis before any claim about clinical substitution.
Frequently asked questions
Was INSL3 measured in serum?
Yes. The study compared it with a local testicular measurement.
Was the setting chronic infertility treatment?
No. It used an acute experimental suppression model in healthy men.
Does strong correlation prove accurate individual prediction?
No. Additional validation is needed.
Can a treatment influence both correlated measurements?
Yes. That possibility should be examined analytically.
Did every candidate marker behave similarly?
No. The reported relationships differed.
Is the study a clinical test-replacement recommendation?
No. It identifies a candidate relationship.
What should follow this appraisal?
Review within-group associations, assay reliability, independent validation, and companion-report overlap.