Hormone stability in human whole blood

Abstract

Objective

To determine whether significant changes in the plasma concentrations of 17 hormones occur when human whole blood is held at 4 or 24°C for up to 24 h before separation of the plasma fraction.

Design and methods

Blood samples (EDTA) from healthy human volunteers were held at 4°C or 24°C for 0.5, 6 or 24 h before separation. Plasma concentrations of ACTH, aldosterone, gonadotrophin α-subunits, AVP, C-peptide, estradiol, FSH, GH, glucagon, IGF-1, IGFBP3, insulin, leptin, LH, prolactin, PTH and VIP were measured and the results compared to baseline values. Nonlinear regression was used to test for a significant mean rate of change. The time interval for median concentrations to change by 10% was determined.

Results

Significant changes were observed for ACTH (decrease at 18.6 hr, 4°C; 17.5 hr, 24°C); AVP (increase at 2.6 h, 24°C); insulin (decrease at 16.8 hr, 4°C; 16.9 hr, 24°C) and VIP (increase at 18.6 h, 24°C). No changes were detected for the remaining analytes.

Conclusions

The measurement of some hormones is compromised by a delay in plasma separation from normal human blood. While many hormones appear stable in normal whole blood, we recommend that processing occurs without delay.

Introduction

Preparation of plasma samples for hormone analysis is not always possible immediately after venesection. The practice in an ideal setting in a specialized endocrine test center is usually the immediate centrifugation of blood specimens at 4°C. In reality, however, many specimens are collected in hospital wards or in external surgeries where a delay before plasma separation from red blood cells occurs frequently. As a referral laboratory specializing in the measurement of human hormones, we have no control over the collection procedures and integrity of the majority of incoming specimens. We are often asked about the suitability of specimens that have had delayed processing for the measurement of particular hormones. While it may be ideal to recommend a re-bleed, this may be inconvenient to the patient. It may even be impossible and have implications for diagnosis and patient management if, for example, a baseline value is required but treatment or surgery has occurred. Information on hormone stability in whole blood has been reported for some of the analytes that we measure [1], [2], [3], [4] although using dissimilar methods of analysis. This study was performed to provide information, relevant to our current hormone testing methods, on restrictions of time and temperature before separation of whole human blood on the subsequent levels of 17 hormones that are measured routinely in our clinical laboratory. We used EDTA plasma as this is the preferred matrix for hormone testing in our laboratory and that in which our reference range values were derived. We also found improved stability, particularly at higher temperatures, for several hormones in EDTA plasma compared to serum aliquots [5]. The end points studied included our recommended practice of separation within 30 min, a 6 h period that represents the time it may take for blood to reach a separation facility from a ward or surgery and a 24 h period that allows for an overnight delay in blood separation. The temperatures were chosen to reflect blood held under refrigeration or in ice/water and samples left at room temperature in a controlled environment such as a laboratory or hospital ward (temperature ranges 22–24°C).