Is fasting necessary for thyroid-stimulating hormone estimation? Role of analytical and biological variation in the interpretation of results

Abstract

Abstract Background: There are varied reports on the effect of sample timing on thyroid-stimulating hormone (TSH) interpretation and its role in under- or over-diagnosing subclinical hypothyroidism. This study was undertaken to evaluate the difference in consecutive values of TSH in fasting and post-prandial samples and interpret the clinical relevance of the difference obtained using parameters such as measurement uncertainty (MU) and reference change value (RCV) which are based on analytical and biological variations. Methods: The fasting and post-prandial TSH were evaluated in 30 paired specimens of apparently healthy individuals. TSH was analysed using chemiluminescence technology. MU was calculated from 6-month cumulative internal quality control data, and RCV was obtained from the published literature. Results: The study showed that fasting TSH levels were higher than post-prandial TSH. However, 90% of the difference observed between the pairs at lower concentrations of TSH was within laboratory MU, and 100% of the difference observed at higher concentrations of TSH was within the MU. When the difference obtained was compared with RCV (which includes both analytical and intra-individual variations), the difference was within the RCV across all concentrations. Conclusions: Although TSH values in the fasting specimens were higher than the post-prandial samples, a comparison of the difference obtained with MU and RCV shows that the difference was well within these parameters and hence less likely to be a clinically significant difference. Thus, the RCV of analytes may be used as a guide in the interpretation of consecutive samples or serial measurements of analytes.