| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
1 Department of Nuclear Medicine, Carl Gustav Carus Medical School, University of Technology Dresden, Dresden, Germany
aAddress correspondence to this author at: Department of Nuclear Medicine, Carl Gustav Carus Medical School, University of Technology Dresden, Fetscherstrasse 74, D-01307 Dresden, Germany. Fax 49-351-458-5347; e-mail Klaus.Zophel{at}uniklinikum-dresden.de.
To the Editor:
In a recent report, Kratzsch et al. (1) reported new reference intervals for thyrotropin [thyroid-stimulating hormone (TSH)] and thyroid hormones, based on National Academy of Clinical Biochemistry (NACB) criteria. Although the report deals with defining reference intervals for TSH and for thyroid hormones, the major point of discussion is the reference interval for TSH. The discrimination of subclinical thyroid dysfunction from euthyroidism has major therapeutic consequences, especially in elderly and/or comorbid patients. With the improved sensitivity of current TSH methodology, TSH measurement has replaced free thyroxine testing as the first-line test for screening and finding thyroid dysfunction. Current 3rd-generation assays (lower limit of detection 
0.01 mIU/L) now allow the detection of subclinical hyperthyroidism (2)(3).
There is consensus regarding the lower limit of the TSH reference interval (0.3–0.4 mIU/L). In contrast, however, the upper limit of the TSH reference interval is currently under discussion. In recent guidelines, the NACB recommended the use of 2.5 mIU/L, rather than 4 mIU/L, because reference populations, on which the definition of the reference interval is based, contain individuals experiencing an initial phase of autoimmune thyroid disease (4), thus skewing the upper reference limit of TSH (5). Ultrasonography, in addition to measurement of thyroid autoantibodies, should be used to exclude these individuals (6)(7). From a whole study group of healthy blood donors (n = 870), Kratzsch et al. (1) defined a constraint group (n = 453) exclusive of individuals with family history of thyroid disease, positive autoantibodies to thyroid peroxidase (TPO) and/or thyroglobulin (Tg), increased free triiodothyronine and/or free thyroxine, and sonographically assessed abnormalities of the thyroid. Interestingly, the upper limit (97.5th percentile) was in the same range for the constraint group as for the whole study population (3.63 mIU/L vs 3.77 mIU/L). Both values were higher than the NACB recommendation (2.5 mIU/L). Our own data (8) confirm these results. Jensen et al.(9) also reported an upper limit of 4.1 mIU/L in 987 healthy volunteers selected from a total population of 1512 persons. They found that 250 individuals had at least 1 thyroid autoantibody, 121 were taking medications other than estrogens and occasional analgesics, and 105 reported a family history of thyroid disease. Furthermore, the 987 healthy adults included both women and men between 17 and 66 years of age. In our study, the 97.5th percentiles were 3.35 mIU/L for the constraint group (n = 713) and 3.34 mIU/L for the whole group (n = 1442; Fig. 1
). As Kratzsch et al. (1) also found in their study, the lower reference limit was higher in our constraint group than in the whole group (0.30 mIU/L vs 0.11 mIU/L). Völzke et al. (10) presented the well-known inverse correlation between TSH and age, as also mentioned by Kratzsch et al. (1) and Jensen et al. (9). According to our data, the upper limits of the disease-free group (n = 1488) and the whole study population (n = 4298) were in the same range. We therefore question whether we really need a reference population selected according to NACB criteria for assessment of TSH reference values.
|
There are several possible explanations for these results. In our opinion, the most important influence is the iodine status of the population investigated. Whereas the National Health Nutrition and Examination Survey (NHANES) III study (11) was carried out in a region of sufficient iodine intake, the European studies were performed in regions with mild or moderate iodine deficiency. Only Jensen et al. (9) compared the iodine intake of individuals from different regions. They showed a lower median and reference interval in persons with mild iodine deficiency compared with those with moderate iodine deficiency. Furthermore, the lower 2.5th percentile in the constraint group of older individuals may have been a result of their longer exposure to iodine deficiency, which could lead to a higher risk of having an initial state of subclinical hyperthyroidism, e.g., disseminated autonomy (12). In contrast to Kratzsch et al. (1), we would favor this reason more than their claim that the lower comorbidity of healthy blood donors would be responsible for the difference.
Several studies indicated that the currently available commercial assays are too insensitive to detect thyroid autoantibodies (TPOAb and TgAb) in an early state of autoimmune thyroid disease (1)(4)(5)(8)(10). Therefore, thyroid ultrasonography is recommended to exclude these individuals. Obviously, however, ultrasonography is not sufficient in populations with lower iodine intake.
A third possible explanation for these findings is that the microheterogeneity of the antigen TSH itself leads to differences in the epitope spectrum. As a result, the interaction of TSH with the assay-specific antibodies differs among assays, or better, among manufacturers (2)(13)(14). The analysis of external quality controls—which is mandatory for each laboratory in Germany—gives insights into this possibility (15). Some assays slightly undermeasure and others overmeasure the control sample consistently, although all manufacturers calibrate their assays to International Reference Preparation (IRP) 80/558. Of course, the intramethod variation is acceptable.
In conclusion, it seems useful to redefine the upper limit of the TSH reference interval to a value lower than 4.0 mIU/L, following the NACB criteria. Despite a family history of thyroid disease, thyroid ultrasonography, and sensitive thyroid autoantibody measurement, the iodine status of the reference population should be known to define a representative TSH reference interval usable for therapeutic decisions, especially in elderly patients. Thyroid-releasing hormone testing might be helpful in central hypothyroidism, as recommended by the NACB (4)(6).
Considering the diversity of problems that impact the establishment of reference intervals for TSH, which were mentioned in the reports of Kratzsch et al. (1), Voelzke et al.(10), and in our own (8), we question whether it makes sense at all to define a common upper limit for TSH determinations.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
