| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
1 Department of Clinical Biochemistry, Odense University Hospital, Odense, Denmark
2 NOKLUS, Norwegian quality, improvement of primary care, laboratories, Division for General Practice, University of Bergen, Bergen, Norway
3 Department of, Endocrinology and Metabolism, Odense University Hospital, Odense, Denmark
aAddress correspondence to this author at: Department of Clinical Biochemistry, Odense University Hospital, DK-5000 Odense C, Denmark. Fax 45-65-41-19-11; e-mail Esther.Jensen{at}ouh.fyns-amt.dk.
To the Editor:
In the current debate on reference interval(s) for serum thyroid-stimulating hormone (TSH) concentrations, a lowering of the upper reference limit from 
4.0 to 2.5, or even 2.0, mU/L has been proposed by the National Academy of Clinical Biochemistry (NABC) (1). This proposal is based in part on the observation that populations with the lowest prevalence of antithyroid antibodies have the lowest TSH upper limits. Other arguments for the lowering of the upper limit of the reference interval are related to the question of whether mild TSH increases have any clinical consequences. This question, however, illustrates the problem of mixing the concepts of (a) decision limits (e.g., discrimination values, cutoffs, action limits), which are based on the clinical consequences and treatment strategies and (b) reference intervals, which are based solely on biology and mathematics applied in an appropriate reference population. This confusion is also addressed in a recent paper on TSH reference interval(s) (2).
The debate for lowering the upper TSH reference limit also includes the argument that the reference distribution for serum TSH should be gaussian in nature, but the upper tail of the distribution is currently skewed by: (a) euthyroid outliers such as may occur in patients recovering from nonthyroidal illness, (b) measurement of bioinactive TSH isoforms, (c) TSH receptor gene polymorphisms, and (d) occult autoimmune thyroid dysfunction. As a consequence, some authorities suggest the distribution tail to be deleted (1). In our opinion, however, this upper tail is an essential part of the distribution. In fact, when all values from individuals at risk are removed, log-gaussian distributions are common for most serum components (3), as we demonstrated for serum TSH, which is unimodal and log-gaussian (4).
We now focus on the newer documentation regarding serum TSH reference intervals and methods. Despite the fact that several publications suggest an upper limit of 4 mU/L, NACB proposes an upper limit of 2.5 mU/L, although only one of several population-based studies supports this (Table 1
).
|
As evidenced in Table 1
, studies vary widely in time of sampling and analytical methods used, as well as inclusion and exclusion criteria. The study with the highest relative median serum TSH in the Deutschen Gesellschaft für KlinischeChemie und Laboratoriumsmedizin hormone survey could have been expected to demonstrate the highest upper reference limit, but clearly it does not (Table 1
), suggesting that factors other than method standardization play a role. Repeated data from external quality assessment performed from 2000 to 2005 disclose that between-method variation is only a minor source of the variation in serum TSH. The exclusion of individuals at risk, however, has been based on nonstandardized criteria, and the importance of time of sampling has been ignored. In fact, in the majority of publications the time of sampling has not been specified.
There is evidence of a considerable diurnal variation in serum TSH concentration, with a maximum around midnight (6). A decrease of up to 50% occurs from 8:00 to 9:30 AM; thereafter the concentration remains relatively constant until evening, with a smaller nadir in the late afternoon. Because serum TSH concentration decreases markedly during the morning and time of sampling is unknown in most studies, sampling time differences between studies may be a primary reason for the discrepancies in published reference intervals. Individuals working night shifts have displaced or reduced diurnal rhythms, a phenomenon that should also be acknowledged (or such individuals excluded) when establishing reference intervals. Consequently, our proposal is to establish reference intervals as a function of time of sampling to reveal the influence of time on reference limits for serum TSH. The outcome of sampling time investigations will indicate whether such data will lead to recommendations for time of sampling or to time-dependent reference intervals.
Studies to establish decision limits for serum TSH should be based on standardized measurements performed in longitudinal follow-up of cohorts with various concentrations of serum TSH. Such studies may well support intervention below a serum TSH concentration of 4.0 mU/L. At present, however, such a decision is not based on unequivocal evidence (2).
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  C. A. Spencer, J. G. Hollowell, M. Kazarosyan, and L. E. Braverman National Health and Nutrition Examination Survey III Thyroid-Stimulating Hormone (TSH)-Thyroperoxidase Antibody Relationships Demonstrate That TSH Upper Reference Limits May Be Skewed by Occult Thyroid Dysfunction J. Clin. Endocrinol. Metab., November 1, 2007; 92(11): 4236 - 4240. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
