⚠️ Full text was unavailable for end-to-end verification. The conclusions below are limited to the abstract and authors’ public analysis repository. gap/no-fulltext-access
Fernandez et al. 2025 — is taurine an aging biomarker?
TL;DR
Across three human cohorts, longitudinal rhesus macaques, and longitudinal mice, circulating taurine increased or remained stable with age. The paper strongly challenges low circulating taurine as a universal aging biomarker, but it did not administer taurine or attempt to replicate Singh’s mouse lifespan, methylation, or healthspan experiments.
Design
- Baltimore Longitudinal Study of Aging: n=742, ages 26–100, repeated samples within participants.
- Balearic Islands Study of Aging: n=72, ages 20–85, cross-sectional.
- Predictive Medicine Research cohort: n=159, ages 20–68, cross-sectional.
- Rhesus macaques: n=32, repeated measurements across approximately 3–32 years of age.
- Study of Longitudinal Aging in Mice: n=39, repeated measurements from 9–27 months.
The public analysis package includes sex-stratified models and sensitivity analyses adjusting the Baltimore cohort for diet, excluding outliers, and restricting to active participants.
Key results
- Circulating taurine increased or remained unchanged with age across the analyzed populations.
- Between-person variation was large relative to the age-associated within-person change.
- Associations with strength, motor function, and energy-homeostasis outcomes varied across cohort, sex, and endpoint.
Interpretation
This is a premise challenge, not a failed intervention replication. It weakens the claim that normal human aging creates a general taurine-deficiency state requiring replacement. It cannot determine whether taurine has pharmacological benefits for a defined disease or phenotype.
Limitations
- No taurine intervention, aging-clock, lifespan, or hard healthspan endpoint.
- Platforms and matrices varied across cohorts, although cross-platform concordance and within-cohort longitudinal results argue against assay differences explaining the overall direction.
- Full text remains inaccessible for an end-to-end verification pass. gap/no-fulltext-access