Establishing reference intervals for galectin-3 concentrations in serum requires careful consideration of its biological determinants

Background: Appropriately established reference intervals for laboratory biomarkers may help the interpretation of their results and facilitate clinical utilization. Objectives: i) To determine reference intervals for serum galectin-3 measured using the Architect STAT immunoassay, and ii) to identify factors affecting galectin-3 concentrations. Methods: We recruited 533 questionnaire-identified apparently healthy individuals, in which laboratory biomarkers were used to detect asymptomatic myocardial injury and dysfunction, ongoing inflammation, hyperglycemia, dyslipidemia, and renal dysfunction. A final reference group of 180 subjects was selected. Results: 2.5th and 97.5th percentiles of distribution of galectin-3 concentrations in the reference group (90% confidence interval) were 5.9 (5.0–6.8) and 18.1 (17.2–19.0) μg/L, respectively. Older age contributed to higher galectin-3 concentrations, but influenced derived reference intervals to a lesser extent. Other major determinants of galectin-3 concentrations observed in the questionnaire-screened population were not linked to galectin-3 in reference individuals. In aiming to decide if reference limits should be partitioned by age, we compared galectin-3 concentrations in subjects < 40 (n = 124) and ≥40 years (n = 56). Higher galectin-3 concentrations were found in older people (median, 12.4 μg/L vs. 11.5 μg/L, p = 0.016). Accounting for manufacturer's declared imprecision at galectin-3 physiological concentrations of < 10% (as CV) and the estimated URL for subjects ≥40 (18.8 μg/L) and < 40 (17.9 μg/L), we recommend the adoption of a single URL for the overall adult population. Conclusions: We established reference intervals for galectin-3 in which the effects of biological determinants were irrelevant. Although in healthy subjects age may affect galectin-3 release, this does not appear to necessitate age-related reference limits.