Interpretation of Results for Thyrotropin In Serum To the Editor: For many years, Welicome Diagnostics (Beckenham, Kent, BR3 3BS, U.K.), which is a division of the Wellcome Foundation, has operated an unmunoassay quality-assessment program for the benefit of laboratories that do hormonal and other immunoassays. Although this scheme is U.K. based, its participants are located worldwide. More than 1000 laboratories now subscribe to the scheme, of which about 130 are in the U.K. However, most do not perform all assays and so the number of returns for each analysis is significantly less. The program consists of two six-month cycles! annum of lyophilized serum specimens, which are reconstituted and assayed every two weeks. In recent cycles, participants have been invited to give a diagnosis associated with their result; for example, they would be asked to state whether they regarded their cortisol result to be low, normal, or above normal. For thyrotropin assay, participants state whether they consider their restilts are compatible with primary hypothyroid, euthyroid, or primary hyperthyroid status. We examined the summary data (see Table 1) for the last full cycle, and noted that about 10% of the laboratories misdiagnose results that are clearly in the primary hypothyroid range. Such misdiagnosis could be a result of laboratory imprecision, inaccuracy, or use of an inappropriate reference interval. We have tried to establish whether these data are methodrelated, but this information is not easily retrievable from the program. However, if any of these reasons were responsible, the incidence of misdiagnosis would be greater when the alllaboratory mean value was equivocal. This is not the case, because the rate of misdiagnosis is greatest when the alllaboratory mean is unequivocally in the primary hypothyroid range. We can only conclude, therefore, that this results from a misunderstanding of the clinical significance of thyrotropin measurement. This in turn gives rise to considerable concern when it is realized that 10% of participating laboratories are making this mistake (we do not know whether it is the same 10% each time), and we wonder whether this is extended to the interpretation of patients’ results. In the light of these observations, it behooves all clinical chemists to consider their own interpretation of resuits for thyrotropin, whether or not their laboratory subscribes to the Wellcome Quality Assessment Programme. A. P. Roelli H. G. J. Worth Clin. Chem. Dept. King’s Mill Hospital Sutton-in-A shfield Nottinghamshire, NG1 7 4JL, U.K. Correlations between Abnormalities in Chromium and Glucose Metabolism in a Group of Diabetics To the Editor: We have previously reported (1) that the chromium (Cr) concentration in randomly chosen plasma samples is lower in diabetics than in controls. We Table 1. Clinical Assessment of Results for Thyrotropin from a Complete Cycle of the Weilcome Quality Assessment Programme MI-laboratory mean result for thyrotropln, mull-mt. unlts/L 1.61 1.45 15.13 2.10 1.59 15.23 1.48 2.11 15.27 1.63 2.11 1.45 Numbsr of laboratorIes makIng each InterpretatIon Primary Primary hypothyroidIsm 3 8 522 26 4 487 9 6 507 Euthyroldlsm hyperthyroldlsm 523 595 3 4 7 53 534 575 12 561 586 14 4 7 543 8 3 3 572 583 540 2.04 496 13 46.83 These data are reproduced by the kindpermission of WeilcomeDiagnostics. 4 3 6 5 5 5 0 42 therefore performed a six-month longitudinal study on diabetics (Type 1, n = 11; Type 2, n = 5) attending an outpatient clinic. At each visit, blood and urine were taken for estimation of Cr, glucose, and creatinine. Each patient provided one 24-h urine specimen during this period. Results were compared with those for a control group of 11 healthy volunteers. For chromium analysis we used an electrothermal atomic absorption technique (2). Glucose was measured by the “Glu-cinet” method (Technicon no. ID/SM 4/001 F83) and creatinine by a modified Jaffe method, in a Technicon RA 1000 analyzer. Results are expressed as mean ± SD. In a preliminary report (3) we showed that, for the whole group (Types 1 and 2), plasma Cr was 5.9 ± 3.5 nmol/L, which was significantly (P <0.001) lower than in the control group (9.2 ± 3.6 nmoLfL); however, 24h Cr excretion was 2.7 ± 1.3 imol per mole of creatinine, which was three times larger (P <0.001) than in the control group (0.95 ± 0.42 zmol per mole of creatinine). Together, these findings imply a primary renal lesion, leading to an increase in Cr excretion. This must (at steady state) be compensated by an equivalent threefold increase in absorption of Cr from the gut, as found previously by use of radiotracer methods (4). This increased absorption is nevertheless unable to restore plasma Cr to normal. In our 11 Type 1 diabetic patients, 24-h Cr excretion was correlated with 24-h glucose excretion (r = 0.65; P <0.05). This suggests that either poor control exacerbates the renal lesion or, when this lesion is worse, good control is more difficult. Chromium excretion also correlates with duration of disease (r = 0.62; P <0.05), suggesting that the renal defect gets worse with time. However, all individual values for plasma Cr concentration (six values for each patient, measured during six months) correlate with individual values for plasma glucose (r = 0.27; P <0.05). We have shown (5) that, in healthy individuals, an oral 75-g glucose load decreases plasma Cr. This is not accounted for by increased renal loss and presumably involves Cr uptake by “stores” (perhaps cells). Therefore, in the present group of patients, it may be that temporary insulin lack (evidenced by high plasma glucose) is associated with a relative shift of Cr out of stores into plasma. Thus small, short-term variations in plasma Cr may be ascribable to internal shifts, even when mean plasma Cr is de- CLINICAL CHEMISTRY, Vol. 34, No. 7, 1988 1525 creased because of increased renal loss of Cr. We conclude that the abnormalities in the metabolism of Cr and glucose in diabetics seem to be related. However, it is not yet clear (a) whether Cr depletion has a causative role in the pathogenesis of diabetes mellitus, as has been shown in patients receiving longterm total parenteral nutrition (6), and (b) whether renal Cr loss develops as a consequence of early diabetic nephropathy. References 1. Morris BW, Kemp GJ, Hardisty CA. Plasma chromium and chromium excretion in diabetics [Letter]. Clin Chem 1985;31:334-5. 1526 2. Morris BW, Kemp GJ. Chromium in plasma and urine measured by electrothermal atomic absorption spectroscopy [Letter]. Clin Chem 1985;31:171-2. 3. Morris BW, Kemp GJ, Hardisty CA. Alterations in plasma and urine chromium in diabetes mellitus [Abstract]. J Endocrinol 1986;108:298. 4. Doisy RJ, Streeton DHP, Souma ML, et al. Metabolism of 51chromium in human subjects, normal, elderly, and diabetic subjects. In: Mertz M, Cornatzer WE, eds. New trace elements in nutrition. New York: Marcel Dekker, 1971:155-68. 5. Morris BW, Grifliths H, Kemp GJ. Effect of glucose loading on concentrations of chromium in plasma and urine in healthy adults. Clin Chem 1988;34:1114-6. 6. Jeejeebhoy KN, Chu RC, Marliss EB, et al. Chromium deficiency, glucose intoler- CLINICALCHEMISTRY, Vol. 34, No. 7, 1988 ance and neuropathy reversed by chromium supplementation in a patient receiving long-term total parenteral nutrition. Am J Clin Nutr 1977;30:531-8. Brian W. Morris Huw Griffiths Graham J. Kemp1 Dept. of Clin. Chem. Northern General Hospital Herries Road Sheffield S5 7A U, U.K. 1 Present address: Dept. of Human Metab. and Clin. Biochem., Med. School, Royal Hallamshire Hosp., Sheffield, U.K.
© Copyright 2026 Paperzz