A blood-based enzyme signal was significantly lower in postmenopausal women with osteoporosis than in age-matched healthy controls. Mean serum XT-I activity was approximately 0.46 plus or minus 0.17 µU/mL in the osteoporosis group, compared with 0.82 plus or minus 0.22 µU/mL in controls. The reported p-value was below 0.0001.
The comparison involved 40 postmenopausal women diagnosed with osteoporosis and 40 age-matched healthy postmenopausal controls. Serum XT-I and XT-II activity was measured with a mass-spectrometry enzymatic assay that tracked xylose incorporation into specific acceptor peptides. The study was cross-sectional, meaning it compared measurements from a defined period rather than following people over time, so it can show associations but not cause and effect.
One isoform stood out
XT-II did not differ significantly between the groups; its reported p-value was at least 0.05. The authors therefore describe XT-I as an isoform-specific parameter associated with the osteoporotic pattern and suggest that it merits further study as a biomarker and possible therapeutic target.
XT-I activity was positively associated with bone mineral density scores, or BMD scores, at the lumbar spine and femoral neck. The reported correlation values, which describe the direction and strength of a relationship, were 0.67 for lumbar-spine T-score, 0.65 for lumbar-spine Z-score, 0.63 for femoral-neck T-score and 0.51 for femoral-neck Z-score. Each reported p-value was below 0.0001.
The pattern ran in the opposite direction for fracture-risk and bone-turnover measures. Higher XT-I activity was associated with lower FRAX hip and major-fracture scores and with lower levels of P1NP and CTX. Correlation values were -0.45 for hip FRAX, -0.38 for major-fracture FRAX, -0.36 for P1NP and -0.32 for CTX. FRAX is a composite estimate that includes clinical and BMD inputs, so these correlations are not independent biological evidence.
The study's wider clinical measures showed the same overall separation between the groups. Compared with controls, women with osteoporosis had lower T- and Z-scores at both measured sites and higher CTX, P1NP and FRAX scores. The BMD and FRAX comparisons had p-values below 0.0001, while the CTX and P1NP comparisons had p-values below 0.01.
A wider statistical pattern
Researchers then used principal component analysis, or PCA, to summarize how the measurements moved together across the cohort. The first PCA axis accounted for 29.6% of the total variance and the second for 12.4%. XT-I and the BMD variables aligned positively with the first axis, while FRAX hip, FRAX major fractures, CTX and P1NP aligned negatively. Osteoporosis participants were mainly on the negative side of that axis and controls toward the positive side, although the groups partly overlapped.
XT-II showed a much narrower set of reported associations. It was negatively associated with age and FRAX scores and positively associated with lumbar-spine T-score, with correlation values of -0.23, -0.24 and 0.25, respectively; each reported p-value was below 0.05. Those links were weaker and less extensive than the XT-I pattern.
What the study cannot answer
The findings do not show that low XT-I causes osteoporosis. The study's cross-sectional design cannot establish whether lower XT-I comes before osteoporosis, results from it or changes alongside it. The authors say clinical validation will require larger prospective cohorts.
The authors also report that direct evidence linking circulating XT-I to bone-specific glycosaminoglycan, or GAG, composition is lacking. The study included only postmenopausal women, so it does not establish whether the same pattern applies to men or premenopausal women.
The control group was an age-adjusted clinically healthy convenience sample rather than a population-based reference cohort. Controls had T-scores greater than -1.0 at both measured sites, and there were no reported bone-metabolism medications or secondary-osteoporosis conditions among them.
Further work will need to establish diagnostic performance, reference intervals and biological variability for XT-I. Longitudinal endocrine measurements and bone-matrix analyses could help determine whether XT-I changes during osteoporosis progression or treatment, and whether the assay adds useful information beyond established bone-turnover markers.
Research transparency
The analyzed datasets are available from the corresponding author upon reasonable request. The authors report no competing interests. The research received DFG grant FA1381/1-2 funding to Isabel Faust-Hinse, with open-access support from Projekt DEAL and Ruhr-University Bochum funds.
Paper data and sources
Original title: Xylosyltransferase-I serum activity is a promising biomarker for postmenopausal osteoporosis.
Authors: Alina Witt, Massimo De Martinis, Matthias Kühle et al.
Journal/Repository: Scientific reports
Status: Peer-reviewed
First online: 2026-08-20
DOI: 10.1038/s41598-026-66889-0
Original paper · Full text