Measurement of steroid hormones in saliva: Effects of sample storage condition

Measurement of steroid hormones in saliva is increasingly common in elite sport settings. However, this environment may enforce handling and storage practices that introduce error in measurement of hormone concentrations. We assessed the influence of storage temperature and duration on reproducibility of salivary steroid levels. Nine healthy adults provided morning and afternoon saliva samples on two separate occasions. Each sample was divided into identical saliva aliquots which were stored long-term (i.e. 28 and 84 days) at - 80°C or - 20°C (testing day 1), and short-term (i.e. 1, 3, 7 and 14 days) at 4°C or 20°C (testing day 2). Samples were analyzed for cortisol, testosterone and estradiol using ELISA. In non-freezer conditions, there was a decrease from baseline to 7 days in testosterone (- 26 ± 15%) and estradiol (- 58 ± 17%) but not cortisol concentrations (p < 0.001). This decrease was larger in samples stored at room temperature than in the refrigerator (p ≤ 0.01). There were small but significant changes in measured concentrations of all hormones after 28 and/or 84 days of storage in freezer conditions (p ≤ 0.01), but these were generally within 12% of baseline concentrations, and may be partly explained by inter-assay variability. Whole saliva samples to be analyzed for cortisol, testosterone and estradiol should be frozen at - 20°C or below within 24 h of collection, and analyzed within 28 days. Storage of samples for measurement of testosterone and estradiol at temperatures above - 20°C can introduce large error variance to measured concentrations.