Fig. 2: Age-dependent characterization of QPLEXTM biomarkers in cognitively normal (CN) participants (2a-2c, CN− only; 2d-2f, entire CN group).
a A monotone spline model for QPLEXTM biomarkers. The ages of the participants acted as a proxy for time. The levels of QPLEXTM biomarkers were transformed to z scores. The patterns of each biomarker revealed differences based on age groups (the 1st tertile, ≤42 years; 2nd tertile, 42–65 years; 3rd tertile, >65 years). The curves were created by a smoothing spline with four knots. b Pearson’s correlation between the QPLEXTM biomarkers. Colored and white boxes show R and P values, respectively. We confirmed that all the variables had low values of variance inflation factors (VIF < 2). c No significant differences in QPLEXTM biomarkers between males and females. P values were obtained from a two-sided unpaired t test. d–f Correlation between the QPLEXTM biomarkers and cerebral amyloid deposition (SUVR) based on age group (d, 1st tertile; e, 2nd tertile; f, 3rd tertile). The levels of QPLEXTM biomarkers were transformed to z scores. Only the 3rd tertile group showed significant results. P values were obtained from Spearman correlation analysis (cutoff, *p < 0.05, and **p < 0.01). Abbreviations: Aβ1–40 beta-amyloid 1–40, POSTN periostin, LGALS3BP (LGAL) galectin-3 binding protein, ACE angiotensin-converting enzyme, SUVR standardized uptake value ratio.
