The reliability of measurements of plasma osmolality is known to be biased by technical artifacts, such as the anticoagulant and the osmometric technique used; the resulting measurement errors therefore may cause errors in interpretation of data. In assessing the potential biasing influence of procedural variables, we found that the temperature at which fresh plasma samples were stored, the duration of storage, and the freezing and thawing of samples appeared to significantly (P < 0.01) affect osmolality values around the narrow physiological range. These factors should be considered in the interpretation of studies on the osmoregulation of vasopressin secretion. In particular, the results suggest that data obtained for any but fresh samples, whether frozen-thawed samples or samples stored at room temperature, are unreliable.
Bohnen, N., Terwel, D., Markerink, M., ten Haaf, J. A., & Jolles, J. (1992). Pitfalls in the measurement of plasma osmolality pertinent to research in vasopressin and water metabolism. Clinical Chemistry, 38(11), 2278-2280. https://doi.org/10.1093/clinchem/38.11.2278