The pre‐analytical stability of 25‐hydroxyvitamin D: Storage and mixing effects

Abstract Background There is an increasing demand for serum 25‐OH VitD testing globally, and this has led to the greater use of automated immunoassays. These may be more prone to non‐specific interference, that is thought to be related to pre‐analytical stability of biological samples. We have investigated the changes in serum 25‐OH VitD concentrations that are caused by storage and mixing conditions, and if such changes are statistical, or clinically important. Methods Blood samples were collected into plain tubes from 31 healthy donors. After separation, serum samples were stored at −20°C and analysis was carried out with and without mixing (vortexing) at different time intervals of days (0, 1, 2, 3, 4, 5, 15, and 30). All samples were analyzed using a chemiluminescent immunoassay. Results Mean serum 25‐OH VitD concentrations for subsequent days of storage compared with day 0 showed a significant time effect (P < .05) except for the samples on day 1 (P = .654) in non‐vortexed samples and day 2 (P = .087), 5 (P = .118) and 30 (P = .118) in vortexed samples. Comparing values for vortexed and non‐vortexed samples on the same day, serum 25‐OH VitD showed a significant difference on days 1 (P = .003), 4 (P = .037), 5 (P = .002), and 30 (P = .025). However, the maximum change value was 8.85% which was less than the known total allowable error (TEa) and reference change value (RCV) for serum 25‐OH VitD. Conclusion 25‐OH VitD is pre‐analytically stable after long‐term sample storage at −20°C and can be analyzed without vortexing. This may be beneficial for both research and diagnostic laboratories.

It is possible that pre-analytical factors may affect the reliability of serum 25-hydroxy vitamin D (25-OH VitD) estimations, and this is important if appropriate clinical interventions are to be made. Hence, it is important to know how pre-analytical factors may affect the results of testing. These factors include sample stability on storage and processing of samples after defrosting. This would be particularly important when analysis is not carried out on a daily basis, and samples are frozen for various lengths of time before estimation. However, to date, there are limited published data on pre-analytical stability of 25-OH VitD in human blood following storage at various temperatures, and for various periods. One study focused on the effects of repeated freeze-thaw cycles. 5 Another study investigated the stability of serum 25-OH VitD levels over time under various sample storage conditions using specimens from only eight patients. 6 A recent study examined the effect of storage at −20°C in the primary collection tube on 25-OH VitD measured by immunoassay. 7 These three studies and other studies (Table 1) have reported vitamin D to be a stable analyte on storage with no concern about using the quality requirement parameters to estimate the clinical decision thresholds.
However, these studies have been limited either by the small numbers of specimens examined or by the chosen time intervals for storage. Furthermore, no previous study has examined the effects of whether samples were mixed following thawing on 25-OH VitD stability. Vortex mixing is usually used to ensure homogenity of sample following thawing of frozen samples, adding an additional step to the analytical process.
Therefore, the aim of the present study was to assess the stability of 25-OH VitD over a longer period of time and to investigate the effect of vortex mixing on the results of serum 25-OH VitD analysis.
In addition to this, our study aimed to investigate whether the potential changes in the results are clinically significant or not, as this has not been discussed in the previous studies.

| Subjects identification
Subjects were recruited randomly for the study from healthy blood donors at King Abdulaziz Medical City, Western Region (National Guard Health Affairs (NGHA)) without regards to vitamin D intake.

| Specimens collection
All specimens were obtained on a single day. Blood was collected from each subject into two plain (red top) tubes (Becton-Dickinson) using standard venesection technique. Samples with insufficient volume (<8 mL of blood), lipemia, or hemolysis were excluded from the study. Specimens were centrifuged within 1 hour of collection, and the serum separated and aliquoted into 1.0 mL Eppendorf tube with 500 µL of serum. Aliquots were stored at −20°C except for two per subject that were analyzed on the day of collection (day 0). One aliquot was vortexed for 10 seconds, and the second analyzed without vortexing. This procedure was repeated using the stored aliquots on days 1, 2, 3, 4, 5, and 30.
All specimens were measured for total 25-OH VitD concentration

| Statistical analysis
Statistical analysis was done using SPSS software version 22.
Comparison between group means was performed using paired t tests and across different groups using analysis of variance (ANOVA). A twotailed t test was conducted using an α level of .05 for all statistical tests.

| RE SULTS
The study comprised 31 subjects, with a mean age ± SD of

| Storage effect
For non-vortex mixed samples, the mean concentration of 25-OH VitD on day 0 was not found to differ significantly on day 1, (P = .654). However, on day 2, the level was significantly higher (P < .001) by 8.50%. This was followed by fluctuations in the mean value on subsequent days, with a slight decrease in the mean on day 3, then a slight increase on day 5, but in all cases, the mean remained significantly higher compared with the mean at day 0 ( Table 2). An

| D ISCUSS I ON
The  This study had the limitation that the RCV was obtained using the LIAISON, Diasorin analyzer, alone. Therefore, it is suggested in the future to obtain the RCV on Abbott, Architect i2000 as this was not provided in previous literatures. Another limitation that the study was designed to include healthy subjects only but those with other pathological conditions with extreme 25-OH VitD values were not considered.

| CON CLUS ION
In conclusion, delayed analysis and running samples without vortexing have no "clinically significant" effect on 25-OH VitD level. Such

G UAR ANTOR
Dr A Borai guarantor for this paper.