Saliva sampling method affects performance of a salivary α-amylase biosensor

Prompted by the discordance between a standardized salivary alpha-amylase (sAA) biosensor applied in clinical settings and a reference laboratory analyzer, we examined the impact of the saliva sampling method on the analytic performance of the biosensor.

Direct mouth readings using the biosensor from 31 normal, healthy volunteers were compared to biosensor and conventional assay readings obtained from saliva samples collected concurrently by passive drool and processed in three different ways (unprocessed, thawed, and thawed and centrifuged).

The direct readings from the mouth showed consistently lower sAA values (Ps < 0.01) compared to all other combinations of sample processing and quantification platforms (biosensor vs. conventional assay). Readings obtained from passive drool saliva were strongly correlated with one another, and Bland-Altman plots of agreement indicated a smaller discrepancy between conventional and biosensor readings obtained with passive drool when compared to the direct mouth readings.

Our study indicates that the analytical performance of the sAA biosensor is influenced significantly by the saliva sampling method. In contrast to the relatively imprecise direct mouth measurements, biosensor sAA levels established indirectly from passive drool saliva samples provide more accurate estimation of sAA levels, even after intermediary processing steps (e.g., freezing, thawing, centrifugation).