Human basic tear fluid osmolality. II. Importance of processing strategy.

Osmolalities of 100 human inferior tear prism fluid samples collected from two subjects and 100 standard solution samples (290 mOsm/Kg) were determined from their melting-point temperatures with the Clifton Nanoliter Osmometer. Accuracy and reliability comparisons were made for endpoints obtained using a single-sample simultaneous-recalibration (SS/SR) strategy, vs a lowest-of-six periodic-recalibration (LS/PR) strategy. Tear fluid osmolality values based on the LS/PR strategy were significantly lower (hypotonic) than for the SS/SR strategy (overall mean = 302.4 and 307.8 mOsm/Kg, respectively; p < 0.0012). The mean difference (5.4 mOsm/Kg) resulted from the combined effects of machine drift and reduced reliability of endpoint determination for tear fluid in comparison to that of a homogenous standard solution. LS/PR osmolality was linearly correlated to SS/SR osmolality with a slope of 0.433 and the two endpoint strategies resulted in equivalent values at 298.4 mOsm/Kg. Eighty-four percent (84%) of individual osmolality readings were greater than 298.4 mOsm/kg, and an 'averaging-type' effect caused LS/PR osmolalities to be increasingly less than SS/SR values above this value. These outcomes from two subjects were approximated by a parallel statistical model. We suggest that physiological representation of basic human tear fluid is more accurate using the single-sample simultaneous-recalibration strategy. Basic human tear prism fluid is more hypertonic and has a greater within-subject range in normal (non-dry eye) humans than previously realized.