Plasma but not salivary p-Tau181 reflects Alzheimer's disease in a Latin American Cohort

Identifying accessible and reliable biomarkers for Alzheimer's disease (AD) remains a major challenge, particularly in low- and middle-income countries. Phosphorylated tau at threonine 181 (p-tau181) measured in plasma has shown strong diagnostic performance, but its potential in saliva, a truly noninvasive biofluid, remains uncertain. This study compared plasma and salivary p-tau181 levels, assessed their agreement, and evaluated their diagnostic accuracy in a Latin American cohort. Eighty participants were clinically classified as cognitively unimpaired (CU, n = 25), mild cognitive impairment (MCI, n = 22), or Alzheimer's dementia (AD, n = 33). Plasma and salivary p-tau181 concentrations were quantified using Single Molecule Array (Simoa) assays. Salivary p-tau181 levels were markedly higher than plasma levels (900.26 vs. 26.67 pg/mL; p < 0.001) but showed no correlation. Bland-Altman analysis revealed a mean bias of -1.56 with significant proportional bias (β = 0.73; p < 0.001), and Passing-Bablok regression confirmed the absence of a linear relationship between matrices. Plasma p-tau181 showed a numerical increase across the cognitive continuum, reaching statistically significant differences only when AD was compared with CU and MCI (AUC = 0.82; 95% CI 0.73-0.92), whereas salivary p-tau181 failed to discriminate clinical groups (AUC = 0.55, ns). These results demonstrate that plasma and salivary p-tau181 are not interchangeable and that current saliva-based quantification methods lack clinical reliability. This study provides the first evidence from Latin America supporting the diagnostic validity of plasma, but not salivary, p-tau181, and highlights the need for further investigation into pre-analytical and biological determinants of salivary biomarker variability.