A broad signal, not a final answer
A study of serum chemistry in calcific aortic valve stenosis, or CAVS, has produced a broad metabolic signal and flagged serotonin for more focused investigation. The work was designed to characterize metabolic alterations associated with CAVS and to assess whether a circulating marker might help detect it. Its central message is cautious: serotonin was chosen as a candidate for follow-up, but the discovery screen did not identify serotonin itself as a statistically significant differential metabolite. The result is therefore a research lead, not an established marker.
The researchers used a two-phase design. First came an untargeted serum metabolomics screen, which surveys the measured chemical profile without limiting the search to one named substance. An independent targeted phase then focused on serotonin. This arrangement allowed the candidate to be selected from the wider screen and assessed in a separate group, rather than being chosen and tested in the same participants.
The discovery cohort initially enrolled 30 people with severe CAVS and 30 age- and sex-matched controls. One participant from each group was excluded because the blood sample did not meet quality requirements. The metabolomics analysis therefore included 29 severe-CAVS participants and 29 controls.
To build the profiles, the researchers used ultra-performance liquid chromatography, or UPLC, coupled with high-resolution tandem mass spectrometry in both positive and negative ion modes. The screen detected 9,777 ions. Its differential analysis identified 707 ions that differed between the groups; 122 remained significant after false-discovery-rate correction. Of the signals that could be annotated as secondary metabolites, 70 were reported, split evenly between 35 upregulated and 35 downregulated metabolites.
Serotonin appeared in a much narrower part of that picture. Only one metabolite matched the serotonin-related pathway. Serotonin itself had an FDR-adjusted q value of 0.48, above the study's threshold of less than 0.05. The researchers selected it for targeted assessment because of pathway evidence, biological plausibility and feasibility, rather than because it had emerged as a directly identified differential metabolite.
A separate test of the candidate
The independent targeted-assessment cohort included 303 participants: 101 controls, 101 people with non-severe CAVS and 101 with severe CAVS. The group structure let the researchers compare the candidate across the control, non-severe and severe CAVS groups. It also kept the candidate assessment distinct from the 29-versus-29 discovery screen.
In that targeted phase, serum serotonin was measured with a commercial ELISA, a laboratory assay aimed at one chosen substance. Samples were coded, laboratory personnel were blinded, and each sample was measured in triplicate. The mean of the three readings was used in the analysis. Those procedures describe how the candidate was measured; they do not turn a hypothesis-generating signal into an established marker.
What the evidence says
The evidence supports a narrower conclusion than the idea of a ready-to-use serotonin test for valve disease. The broad screen found many metabolic differences, while the serotonin-related signal was sparse and did not make serotonin itself a discovery-phase hit. The study supplies a reason to investigate serotonin in a separate targeted analysis, not a basis for treating it as an established standalone marker.
That sequence is the important part of the result. A wide search identified a pattern of differences, then a sparse serotonin-related clue led to a focused measurement in an independent cohort. The study therefore leaves serotonin in the category of a candidate that needs testing, rather than a marker established by the discovery screen itself.
Paper data and sources
Original title: Circulating serotonin is associated with calcific aortic valve stenosis: a two-phase metabolomics study
Authors: Gao W, Guo C, Sarrafzadegan N et al.
Journal/Repository: Frontiers in pharmacology
Status: Peer-reviewed
First online: 2026-10-08
DOI: Not available
Original paper · Full text