Establishment of Reference Ranges for Thyrotropin ...

Wiedemann et al.: Reference ranges for thyrotropin and thyroid hormones in paediatrics 277

Eur. J. Clin. Chem. Clin. Biochem.Vol. 31, 1993, pp. 277-288© 1993 Walter de Gruyter & Co.

Berlin · "New York

Establishment of Reference Ranges for Thyrotropin, Triiodothyronine,Thyroxine and Free Thyroxine in Neonates, Infants, Childrenand Adolescents

By Gudrun Wiedemann, L. Joneiz-Mentzel and Rosel Panse

Klinisch-Chemisches Labor der Klinik und Poliklinik ir Kindermedizin der Medizinischen Hochschule, Erfurt,Germany

(Received November 26, 1992/January 7, 1993)

Summary: Thyrotropin, triiodothyronine, thyroxine and free thyroxine were determined in the sera of 714euthyreotic neonates, infants, children and adolescents (age range 5 days to 18 years), using the IMx fromAbbott Laboratories. Thyrotropin, triiodothyronine and free thyroxine were determined with microparticleenzyme immunoassays (MEIA). Thyroxine was determined with a fluorescence polarization immunoassay(FPIA). The proband collective was divided into 9 age groups, and each age group into males and females.In accordance with the recommendations of the International Federation of Clinical Chemistry, the 95%scatter range was taken as the reference range. In the different age groups, the ranges of some hormonesshowed significant differences between males and females. When no significant difference existed between thesexes, the results for males and females were evaluated statistically as a single group. Only a few referencegroups showed normal Gaussian distributions. Therefore, in addition to the 50th percentile, the 2.5th and97.5th percentiles were also calculated for all reference groups. Minimal and maximal values were alsodetermined. The U-test of Mann & Whitiiey was used to test for significant differences between individualreference groups, and groups showing no significant differences were combined. The corresponding referencesranges were then calculated.

Introduction f . .of free thyroxine in serum under all conditions, be-The concentrations of thyrotropin and thyroid hor- cause of possible interference by albumin, free fattymones in serum are indices of the degree of hyper- or acids and thyroxine-binding globulin (thyropexin)hypoactivity of the thyroid gland. In known distur- 1 — 5).bances of thyroid function, the concentrations of thy- _, . - , . . .., * . " * ι ι r The aim of the investigation was:roid hormones in serum serve as control values for °the efficacy of thyrostatic treatment, or of substitution 1) to establish reference ranges for thyrotropin, triio-therapy with thyroid hormones. Serum thyroxine dis- dothyronine (T3), thyroxine (T4) and free thyroxineplays strong protein binding, so that its concentration (fT4) in healthy neonates, infants, children and ado-is easily affected by the concentration or binding lescents;capacity of its binding proteins. The concentration of ^N ,, . ._ .„ . , ,. .free thyroxine, i. e. the only part of the thyroxine that 2) tO test for S1gnificant sex differences in the studiedis metabolically active, is diagnostically more useful <luantlties within the a§e §rouP^than the total thyroxine concentration. However, 3) to test for significant differences in the studiedthere is still no reliable method for the determination quantities between the age groups.

Eur. J. din. Chem. Clin. Biochem. / Vol. 31,1993 / No. 5

278 Wiedemann et al.: Reference ranges for thyrotropin and thyroid hormones in paediatrics

Materials and MethodsThyrotropin, triiodothyronine, thyroxine and free thyroxinewere determined in the sera of 714 healthy neonates, infants,children and adolescents in the age range 5 days to 18 years.The age composition of the proband collective is summarizedin table 1. Individuals were included or excluded according tothe exclusion criteria of Witt & Trendelenburg (6), which permitthe assembly of a reliable random reference sample at a justi-fiable expense. Only those neonates with a birth weight between2500 and 4000 g and a full term gestation time between 37 and40 weeks were admitted to the 5-day-old age group. Neonateswith hyperbilirubinaemia were excluded, as well as those bornto mothers with acute or chronic illnesses. All probands werefree from any signs of hypo- or hyperthyreosis.

Tab. 1. Age structure of the proband collective for the deter-mination of reference ranges for thyrotropin, triiodoth-yronine, thyroxine and free thyroxine in neonates, in-fants, children and adolescents

Group Age n

Thyro- T3tropin

1 c?1?1

2c?222

3c?323

4<J4?4

5i525

6<f6?6

7cJ7?7

8c?828

9cJ929

5 days5 days5 days2 -12 months2—12 months

2 -12 months2 — 3 years2 — 3 years2 — 3 years4 — 6 years4 — 6 years4 — 6 years7 — 9 years7- 9 years7 — 9 years

10—11 years10—11 years10-11 years12—13 years12-13 years12- 13 years14- 15 years14- 15 years14—15 years16-18 years·16—18 years16—18 years

7173

1431112231819374326694641874555

100464692413778383775

7170

141

1513281818364325684641

874554

99464692413879383876

T4

7071

141

131225

1719364326

694641

874555

100454792403777383573

fT4

5653

109151227

161935

432265453584445397454590393877373774

Test materialAbout 2 ml of blood were taken, with the informed consent ofthe parents and the consent of the Ethics Commission of theMedical University Erfurt, between 08.00 and 10.00 am, froman arm or skull vein, using safety monovettes from the firm ofSarstedt, Numbrecht. Blood samples were centrifuged imme-diately for 5 min at 3000 min"1. The serum was removed witha pipette, then frozen at -22 °C until analysed.

MethodThyrotropin, triiodothyronine, thyroxine and free thyroxinewere all determined with the aid of the IMx from Abbottlaboratories. Microparticle enzyme immunoassays were usedfor thyrotropin, T3 and fT4. The thyiptropin standard 80/588from the WHO was used as a calibrator for the thyrotropindetermination. T4 was determined with the IMx T4 test fromAbbott Laboratories, which is a fluorescence polarization im-munoass y (FPIA). The lower detection limits and cross reac-tivities of the tests are shown in table 2.

Tab. 2. Data provided by the manufacturer on the lower de-tection limits and cross reactivities of the tests used onthe IMx (the concentration of each test substance usedis shown in brackets). The lower detection limit isdefined as the lowest concentration that can be differ-entiated from 0, i.e. the value that is two standarddeviations higher than the zero calibrator.

Test Lowerdetectionlimit

Cross reactivity

IMx hTSHUltrasensitive

0.03 mU/1 with lutropin (3000 U/l):3.3 · ΙΟ'6with follitropin (3000 U/l):5.1 - ΙΟ"4

with human ch rionicgonadotropin (1.7 · 106 U/l):2.2 · 10-8

IMx Total T3 0.15 μg/l

IMxT4

IMx free T4

10.0 μg/l4.00 ng/1

withT4(1100μg/l):none

with T3: < 10%with T3: < 0.5%

Quality control

For the determination of precision from day to day, standardsof low, intermediate and high concentration (from Abbott andfrom Rolf Greiner Biochemica, Flacht) were included intermit-tently in each series. As a measure of the relative methodicalerror, the arithmetic mean (x), standard deviation (s) and thevariation coefficient (CV) were calculated from the individualresults of these control series. Precision is series was monitoredonce, using the control sera, Serodos and Serodos Plus, fromGreiner. Again, the arithmetic mean (x), standard deviation (s)and the variation coefficient (CV) were calculated from theindividual results.

Statistical evaluation of the results

The results were first presented as separate histograms for eachage group and for each sex. Obvious outliers were eliminated.The type of distribution was determined with the Kolmogorov-Smirnov test. If the resulting probability error was below thestated value of α = 0.05, the distribution was assumed to benormal. If the distribution was not normal, the 2.5th, 50th and97.5th percentiles were determined for that age group (7).

Each age group was tested for a significant difference betweenthe sexes, using the U-test of Mann & Whitney. In the absenceof a significant sex difference, males and females were subse-quently treated as a single group. The significance of differences„between age groups was also tested with the U-test of Mann &Whitney.

The degree of any linear relationsKip between age and themeasured quantities was determined by calculation of the cor-relation coefficient^ r.

Eur. L Clin. Chein. Clin. Biochem. / Vol. 31,1993 / No. 5


Results

Thyrotropin

Thyrotropin was determined in the sera of 704 euthy-reotic probands (359 boys, 345 girls). Figure 1 givesan overview of the results for all groups before theywere tested for significance. Although 12 to 13-yearold girls appear to constitute a special group, theywere not significantly different from the 10 to 11-yearold girls (p = 0.6379); they were therefore combinedwith the latter to form a single group. Similarly, othergroups were combined, if they displayed no significantdifference. For each group combination made in thisway the median and the reference range for serumthyrotropin were recalculated; the results are sum-marized in table 3.

Triiodothyronine

Triiodothyronine was determined in the sera of 706euthyreotic probands (363 boys, 343 girls). Figure 2gives an overview of the results for all groups beforethey were tested for significance. Although no signif-icant differences were found between the results for5-day-old neonates, 14—15-year old girls and Ιο-ί 8-year old boys, these groups were treated separately.The new groupings formed after testing for signifi-cance, together with their newly calculated medianvalues and reference ranges are shown in table 4.

Thyroxine

Thyroxine was determined in the sera of 700 euthy-reotic probands (357 boys, 343 girls). Figure 3 givesan overview of the results for all groups before theywere tested for significance. New grouping formedafter testing for significance, together with their newlycalculated median values and reference ranges areshown in table 5.

Fig. 1. Histograms of the 50th percentiles (malesfemales ISS2S3) and the 95% intervals of the serumconcentrations of thyrotropin (mU/1) in the age groups1-9 (see table 1).

t 2οΌΟ;Ξ

4 5 6Age group

Fig. 2. Histograms of the 50th percentiles (males Υ//////Λ,females rssxsNssa) and the 95% intervals of the serumconcentrations of T3 ^g/l) in the age groups 1—9 (seetable 1).

Tab. 3. Summary of results for the serum concentrations of thyrotropin in euthyreotic neonates, infants, children and adolescents,showing 50th percentiles, 95% intervals and the minimal and maximal values. Values are given in mU/1.

Age

5 days2 months— 9 years

10-13 years14—15 years10-13 years16-18 years16- 18 years

Sex

m<?/? 1c? \c?/? J?c?

9

n

143

385

1013837

Median(50thpercentile)

2.17

1.33

1.58

0.88

1.10

Reference range(95% scatter range:2.5th -97.5th percentiles)

0.51-7.90

0.41-3.45

0.64-4.76

0.28-1.88

0.53-2.85

Minimum

0.38

0.12

0.59

0.28

0.53

Maximum

10.6

6.35

9.47

1.88

2.85

Bur. J. Clin. Chem. Clin. Biochem. /Vol. 31,1993 / No. 5

280 Wicdemann et al.: Reference ranges for thyrotropin and thyroid hormones in paediatrics

Tab. 4. Summary of results for the serum concentrations of T3 in euthyreotic neonates, infants, children and adolescents, showing50th percenliles, 95% intervals and the minimal and maximal values. Values are given in μg/l.

Age

5 days

2 months— 3 years

4—13 years

14—15 years

14—15 years16—18 years

16-18 years

Sex

<?/?

c?/?

c?/?

S

1 }9

n

141

6434641

76

38


1.762.292.091.99

1.77

1.56


0.99-3.04

1.30-3.23

1.50-2.70

1.53-2.73

1.16-2.35

1.16-2.30

Minimum

ι

0.711.12

1.18

1.53

0.98

1.16

Maximum

5.21

3.34

5.502.73

2.36

3.20

Tab. 5. Summary of results for the serum concentrations of T4 in euthyreotic neonates, infants, children and adolescents, showing50th percentiles, 95% intervals and the minimal and maximal values. Values are given in

Age

5 days

2-12 months2 — 3 years

2 - 3 years4 — 6 years

10—13 years

7 - 9 years

14-18 years

Sex

c?/?

? J

* 1<f/9 ><?/? J<?/?a/?

n

141

44

278

87150


161

95.5

79.6

88.2

70.6


81.3-233

66.5-158

55.1-113

54.3-130

42.3- 99.1

Minimum

57.8

65.1

46.8

54.0

34.4

Maximum

140

158

146

139140

Free thyroxine

Free thyroxine was determined in the sera of 658healthy probands (340 boys, 318 girls). Figure 4 givesan overview of the results for all groups before theywere tested for significance. There were no significantsex differences in any age group. A weak, significantdifference was found between age groups 7 and 8, but

no significant differences were detectable between agegroups 7 and 9, or between age groups 8 and 9. Thesethree groups also displayed very similar median valuesand reference ranges; they were therefore combined,and the median value and reference range were re-calculated. The new groupings formed after testingfor significance are shown in table 6.

4 5Age group

Fig. 3. Histograms of the 50th percentiles (males W////A,females i SSftQ and the 95% intervals of the serumconcentrations of T« ^g/l) in the age groups 1-9 (seetable 1).

4 5 6Age group

Fig.4. Histograms of the 50th percentiles (males ,females jftySsNftQ and the35% intervals of the serumconcentrations of fT4 (ng/1) in the age groups 1 -9 (seetable 1).

Eur. J. Clin. Chem. Clin. Biochem. / Vol. 31,1993 / No. 5

Wiedemann et al.: Reference ranges for tbyrotropin and thyroid hormones in paediatrics 281

Tab. 6. Summary of results for the serum concentrations of fT4 in euthyreotic neonates, infants, children and adolescents, showing50th percentiles, 95% intervals and the minimal and maximal values. Values are given in ng/1.

Age Sex

5 days cf/?2 months— 1 1 years <J/$

12- 18 years <£/?

η Median(50thpercentile)

109 21.9

308 12.8

241 12.5


13.0-32.9

10.7-17.5

9.2-15.9

Minimum

11.9

9.7

8.00

Maximum

33.4

18.8

18.0

Correlation analysis

Correlations with proband age were sought for theserum concentrations of thyrotropin, T3, T4 and fT4.The serum concentrations of all four hormones de-creased significantly (p < 0.001) with the increasingage (in months) of the studied children: thyrotropin(r = -0.1989), triiodothyronine (r = —0.1259), thy-roxine (r = —0.6540) and free thyroxine (r =-0.5999).

Quali ty control

Results of the quality control are shown in tables 7and 8. With one exception, the variation coefficientswithin series and between series were less than 10%.

Tab. 7. Results for the control of precision from day to day

Thyro-tropin(mU/1)

T3frg/O

T4fog/O

nr4(ng/1)

Control serum

SerodosSerodos PlusAbbott ΜAbbott Η

Abbott LAbbott MAbbott H



Tab. 8. Results for the control

Thyro-tropin(mU/1)

T3(μβ/Ι)

T4(W/l)

fT4(ng/1)

Control serum

SerodosSerodos Plus

SerodosSerodos Plus

SerodosSerodos Plus

SerodosSerodos Plus

n

19315622

223222

263224

131610

X

1.0014.05.84

48.0

0.731.503.75

48.081.3

151

7.1012.533.7

of precision in

n

2222

2323

2020

2323

X

1.0534.4

4.095.89

153190

39.745.6

s

0.080.760.264.24

0.080.140.21

4.405.108.40

0.400.601.40

series

s

0.041.07

0.170.23

5.304.5

1.601.90

CV (%)

8.005.444.458.82

10.99.335.60

9.176.275.56

5.634.804.15

CV (%)

3.723.11

4.153.86

3.492.36

3.994.20

Discussion

In addition to the anamnesis and the interpretationof clinical symptoms, the determination of hormoneconcentrations in serum is extremely important in thediagnosis of pathological endocrine conditions. Toenable the correct evaluation of the serum concentra-tion of a hormone, reference ranges must be estab-lished.

In the present study, reference ranges were determinedfor the serum concentrations of thyrotropin, triio-dothyronine, thyroxine and free thyroxine, using theIMx from Abbott Laboratories. The analytical meth-ods were microparticle enzyme immunoassays(MEIA) and a fluorescence polarization enzyme im-munoassay (FPIA). Both methods produce rapid re-sults, require only a small sample volume, and do notemploy radioactive isotopes. Reference ranges forchildren, using these methods, have not been reportedin the literature. In fact, for most modern immuno-metric assays, the majority of reported referenceranges are for adults. Due to the processes of childgrowth and development, however, clinical chemicalquantities may differ in children and adults. Suchdifferences must be taken into account in the evalu-ation of reference ranges.

In accordance with the recommendations of the In-ternational Federation of Clinical Chemistry, the 95%scatter range (i.e. the range between the 2.5th and97.5th percentiles) was taken as the reference range(7).Each of the four measured quantities in each of thenine age groups shown in table 1 were tested for sex-specific differences, using the U-test of Mann & Wliit-ney. Age groups showing no significant differencesbetween male and female were treated as a singlereference group. The distribution in each group wasinvestigated with the aid of the Kolmogorov-Smirnovtest. In the majority of the age groups, the referencevalues did not show a normal distribution. The ref-erence range was therefore established by reportingthe median value, together with the 2.5th and 97.5thpercentiles. The U-test of Mann & Whitney was also

Eur. J. Clin. Chem, Clin. Biochem. / Vol. 31,1993 / No. 5


used to test for significant differences between theserum hormone concentrations of different agegroups. Age groups showing no significant differenceswith respect to a particular hormone, were generallycombined, and the median value and percentiles re-calculated for that hormone.

To facilitate comparison of the present results withthose from the literature, the latter are presented intables 9 to 12.

The reference ranges for the serum concentrations ofthyrotropin, triiodothyronine, thyroxine and free thy-roxine reported in the present study are not compa-rable with the normal ranges reported by other au-thors, which are shown in tables 9 to 12. This isbecause:

1) different methods were used;

2) the age groups were different;

3) the number of probands in each reference groupwas very small; r

4) with the exception of Borkenstein et al. (8) andStruckmeyer & Raid (9), the authors gave no data onthe type of distribution of the reference values;

5) no details are given about the international refer-ence preparation (WHO standard) that was used forcalibration of the results.Liappis et al. (10—12) gave the reference range asχ ± 2s, and tested for significance between agegroups with Student's t-test. This procedure is validonly for normal Gaussian distributions.

Tab. 9. Reference ranges reported in the literature for the serum concentration of thyrotropin.

Author Method Age group No. Type of distributionof probands and chosen scatter

range

Reference range

Liappis et al., 1988OD

Liappis et al., 1991(12)

LEI A 1-2 days 34(mU/1)

3 -30 days 67

1-12 months 56

1-7 years 68

7 -13 years 77

13-18 years 41

LEI A 1-2 days 15(mU/1)

3 -30 days 46

1-12 months 62

1-7 years 50

7 -13 years 52.,

13-18 years 48

No clear data 9.53 ± 6.797.46to 23.11

1.98 + 1.011.71to 4.00

χ ± Is 1.75 ± 0.95Median value 1.5695% Scatter range to 3.65

1.61 + 0.561.480.50 - 0.73

1.66 ± 0.651.530.36 - 2.96

1.61 + 0.791.390.03 - 3.19

No clear data 8.17 + 5.446.10to 19.05

1.96 ± 1.231.55to 4.42

x ± Is 1.65 ± 0.76Median value 1.5895% Scatter range 0.13 - 3.17

1.47 ± 0.631.440.21 - 2.73

1.42 + 0.521.320.38 - 2.46

- i 1.36 ± 0.561.330.24 - 2.48

Eur. J. Clin, Chem. Clin. Biochem. / Vol. 31,1993 / No. 5


Tab. 9. Continued

Author

Wiedemann et al., 1992(14)

Schönberg & Klett* 1980in Schönberg, 1990(13)

Borkenstein et al., 1980(8)

Struckmeyer & Haid, 1986(9)

Method

MEIA(mU/1)IRP80/558

RIA(mU/1)

RIA(mU/1)

RIA(mU/1)

MRC68/30

Age group No.of probands

5 days 1712

0.5- 1 month 251.5- 2 months 192.5- 3 months 123.5- 4 months 114.5- 5 months 115.5- 7 months 78.0- 12 months 7

2 months— 14 years 182

cJ 4 days2 weeks— 1 year

1 — 6 years6—10 years

10 — 16 years

? 4 days2 weeks— 1 year


10—16 years

Type of distributionand chosen scatterrange

No normaldistribution

Median value95% Scatter range(2.5th -9.5th percentile)

No data

Median value± 1 standarddeviation

No normaldistribution

Median value± 1 standarddeviation

No clear dataMedian value90% Scatter range(5% to 95% quantile)

Reference range

1.9(0.40-9.05)

0.93 ± 0.981.34 + 1.160.99 + 0.831.08 ± 0.960.75 -f 0.860.45 -f- 0.740.77 ± 0.65

2.25 -f 1.96(0.00-6.16)*

4.60 (1.89-9.08)2.40(0.24-5.31)1.80(0.80-6.44)1.95(1-05-3.51)2.65 (0.14-6.05)

4.60 (1.89-9.08)1.80(0.30-5.30)1.75(0.20-6.05)1.85 (1.00-3.36)3.00(0.57-5.83)

* The reference range was taken as the interval between (x — Is) and (x + 2s). Accordingly, the lower limit can be calculatedto be 0.29 mU/1, but the original paper quotes a lower limit of 0.0 mU/l.

Tab. 10. Reference ranges reported in the literature for the serum concentration of tiiodothyronine


range

Reference range

Liappiset al., 1988(ID

LEI A 1-2 days

3 -30 days

1-12 months

1—7 years

34

67

56

68

77

No clear data

± IsMedian value95% Scatter range

7-13 years

13-18 years 41

Eur. L Clin. Chem, Clin, Biochem. / Vol. 31,1993 / No. 5

1.77 ± 0.571.710.63 - 2.91

1.44 ± 0.341.380.76 - 2.12

1.56 ± 0.301.530.96 - 2.16

1.39 ± 0.251.350.89 - 1.89

1.30 ± 0.221.280.86 - 1.74

1.25 ± 0.311.190.63 - 1.87

Continued p. 284


Tab. 10. Continued

Author Method

LiappiseiaL, 1991 LEI A(12) fag/l)

1 »

Sch nberg & Klett, 1980 RIAin Sch nberg, 1990 ^g/l)(13)

Borkenstein et al., 1980 RIA(8) (μ8/ΐ)Struckmeyer & Η aid, 1986 RIA(9) (Mg/1)

Age group

1 -2 days

3 -30 days

1—12 months

1—7 years

7-13 years

13 -18 years

0.5— 1 month1 . 5 — 2 months2.5- 3 months3.5— 4 months4.5— 5 months5.5— 7 months8.0 -12 months

2 months— 14 years

<J 2 weeks3 weeks— 1 year


10- 16 years

S 2 weeks2 weeks— 1 year


10- 16 years

No.of probands

15

46

62

50

52

48

251912111177

182


No clear data

χ ± IsMedian value95% Scatter range

No data

Mean value ± Is

No normal distributionMean value ± Is

No clear dataNo normal distribution

Median value90% Scatter range(5% to 95% quantile)

Reference range

1.71 ± 0.451.710.81 - 2.61

1.38 ± 0.321.400.74 - 2.02

L74 + 0.301.731.14 - 2.34

1.61 + 0.201.651.21 - 2.01

1.55 ± 0.211.541.13 - 1.97

1.43 ± 0.201.381.03 - 1.83

1.98 ± 0.521.84 + 0.542.23 ± 0.471.98 + 0.461.99 ± 0.352.08 + 0.581.81 ± 0.39

1.07 ± 0.34(0.74 - 1.74)*

1.61 (0.94-2.13)1.72(1.22-2.48)1.65(1.18-2.22)1.65(1.28-2.04)1.51 (1.05-^2.11)

1.61 (0.94-2.13)1.79(1.18-2.47)1.65(1.16-2.26)1.49(1.19-2.22)1.41 (0.97-2.23)

* The reference range was taken as the interval between (x - Is) and (x + 2s).

The results of Klett, quoted in Sch nberg (13), mustalso be evaluated critically, since a normal Gaussiandistribution cannot be assumed for such a small num-ber of probands, and it is therefore invalid to calculatethe reference range as x ± 1 s.

The same criticism applies to the work of Borkensteinet al. (8), which was not primarily concerned with theestablishment of reference ranges, but with the de-scription of the normal course of the serum thyroli-berin concentration in childhood. The concentrationsgiven in tables 9, 10 and 11 are the basal values of

the test. Although the authors stated that their dis-tribution of concentrations displayed a skew to theright, they calculated the mean value and standarddeviation for each group, then gave the normal rangeas the interval of (x - 1 s) to (x + 2 s). Since theyfound no significant differences between the selectedage groups, they combined all groups and quoted acommon reference range for the age range 2 months

' to 14 years. In view of the small number of probandsin each age group, the absencp of a normal distribu-tion, and the F-test that was used, this method ofpreceding must be viewed critically.



Tab. 11. Reference ranges reported in the

Author Method

Liappis et al., 1988 LEIA(11) (μΕ/0

Liappis et al., 1991 LEIA(12) fog/I)

Sch nberg & Kielt, 1980 RIAin Sch nberg, 1990 fag/l)(13)

Borkenstein et al., 1980 RIA(8) *g/I)

Struckmeyer & Haid, 1986 RIA(9) feg/l)

.

literature for the serum

Age group

1—2 days

3-30 days

1 — 12 months

1—7 years

7-13 years

13-18 years

1-2 days

3—30 days

1 — 1 2 months

1-7 days

7-13 years

13-18 years

0.5- 1 month1 . 5 — 2 months2.5— 3 months3.5- 4 months4.5— 5 months5.5— 7 months8.0- 12 months

2 months— 14 years

<? 1-3 days

1 week — 1 month

1 month— 1 year

1—6 years

6—10 years

4 4 f\f\** / ·ν τ j»

concentration

No.of probands

34

67

56

68

77

41

28

48

37

65

76

42

251912111177

182

of thyroxine


No clear data


No clear data


No data

Mean value ± Is

No normal distributionMean value ± Is

No clear dataNo normal distribution

Median value

90% Scatter range

*

Reference range

182.1 + 37.7178.5106.6 - 257.5

137.8 + 29.713578.4 - 197.2

95.8 + 21.095.053.9 - 137.8

87.9 + 17.790.052.5 - 123.4

85.5 + 12.88659.9 - 111.1

77.9 ± 14.377.049.2 - 106.6

180.5 + 32.1178.0168.0 - 192.9

158.0 + 40.3164.5146.3 - 169.7

95.0 + 16.796.089.4 - 100.6

89.7 + 17.189.085.5 - 94.0

84.6 ± 14.183.081.4 - 87.9

82.7 + 13.085.578.7 - 86.8

111.1 ± 29.594.7 + 27.192.6 + 21.997.9 ± 18.988.4 ± 24.888.2 + 26.275.6 ± 6.5

80.5 + 20.1(60.4 - 120.7)*

152.0(105.5-194.7)121.0(89.2-186.5)97.0

(74.0-145.0)83.0

(63.9-119.0)90.0

(79.0-118.5)Continued p. 286


Tab. 11. Continued


range

10-16 years (5% to 95% quantile)

$1-3 days

1 week— 1 month

1 month— 1 year

1—6 years

6 — 10 years

10—16 years

Reference range

87.0(77.5-114.5)

152.0(105.5-194.7)121.0(89.2-186.4)108.0(71.4-129.7)9.20

(59.0-108.0)91.0

(63.5-116.0)85.0

(66.3-115.5)

* The reference range was taken as the interval between (x — Is) and (x + 2s).

Tab. 12. Reference ranges reported in the literature for the serum concentration of free thyroxine

Author



Method Age group

RIA 1-2 days(ng/1)

3-30 days

1—12 months

1 — 7 years

7-13 years

13-18 years

LEI A 1-2 days(ng/1)

3-30 days

1-12 months

1 —7 years

7-13 years

No. Type of distributionof probands and chosen scatter

range

12 No clear data

59

50 χ ± IsMedian value95% Scatter range

79

90

51

28 No clear data

48

37 χ ± IsMedian value95% Scatter range

65

76

Reference range

18.08 + 2.3817.7516.57 - 19.60

17.90 ± 19.6018.0017.18 - 18.62

15.82 + 2.3715.2815.15 - 16.49

16.01 + 1.7015.8015.63 - 16.39

16.67 ± 2.1716.2416.22 - 17.13

16.79 + 2.6615.7016.04 - 17.53

26.17 + 5.8525.2016.4 - 37.87

22.28 + 3.6222.5515.04 - 29.52

14.22 ± 1.8613.8010.50 - 17.94

12.09 + 1.9113.009.27 - 16.91

13.38 + 1.9513.109.48 - 17.28



Tab. 12. Continued

Author Method

Sch nberg & Klei t, 1980 RIAin Sch nberg, 1990 (ng/1)(13)

Age group

13-18 years

0.5- 1 month1.5— 2 months2.5- 3 months3.5— 4 months4.5— 5 months5.5- 7 months8.0 — 12 months

No. Type of distribution Reference rangeof probands and chosen scatter

range

42

251912111177

13.68 +13.908.94 -

No data 23.4 +19.6 ±

Median value ± Is 19.2 ±20.2 +17.6 +19.2 ±15.0 ±

2.37

18.42

4.95.54.65.06.75.44.8

In the present study, we followed the recommenda-tions for hypothyreosis screening put forward by theworking group for paediatric endocrinology (13), i. e.that the serum thyrotropin concentration of neonatesbe determined on the fifth day of life. Table 9 showsthe results of an earlier study (14), in which the serumthyrotropin of 1712 healthy neonates was determinedunder the same conditions with the same test on thefifth day of life. The median value is somewhat lowerthan that reported in the present work, and the 95%scatter range is somewhat wider. According to theMann & Whitney test, the slight difference betweenthe two random samples is not significant(p = 0.095), with an error probability of α = 0.05. Incases of suspected congenital hypothyreosis, serumthyrotropin concentrations greater than 10 mU/1should be checked by a control analysis on the fol-lowing day.

The other authors mentioned in tables 9 to 12 inves-tigated neonates both earlier and later than 5 dayspost partum. At the same time, they selected wide ageintervals in these early years, during which relativelylarge age-dependent alterations in the concentrationof this hormone are to be expected. .

With respect to the effect of age on the serum con-centrations of all hormoneSj there is general agree-ment between the results reported in tables 9 to 12and the present results.

In all the investigations, the serum thyrotropin con-centration in neonates was markedly higher than inall other age groups. Thereafter, the serum thyrotro-pin concentration showed a tendency to decreaseslightly with increasing age. This trend was also con-firmed by a correlation analysis.

Tab. 13. Reference ranges for thyrotropin, triidothyronine, thy-roxine and free thyroxine in neonates, infants, childrenand adolescents

Thyrotropin

Males:

Females:

5 days2 months— 15 years16—18 years5 days2 months— 9 years10-13 years14—15 years16-18 years

0.51- 7.90 mU/10.41- 3.45 mU/10.28-1. 88 mU/10.51- 7.90 mU/10.41- 3.45 mU/10.64-4.76 mU/10.41- 3.45 mU/10.53- 2.85 mU/1

Triiodothyronine

Males:

Females:

5 days2 months— 3 years4—15 years

16—18 years5 days2 months— 3 years4-13 years

14-^18 years

0.99- 3.04 μβ/11.30-3.23 μg/l1.50-2.73 μg/l1.16-2.35 μg/l0.99- 3.04 μg/l1.3 -3.23 μg/l1.50- 2.70 μg/l1.16-2^g/l

Thyroxine

Males:

Females:

5 days2—12 months2 — 6 years7 — 9 years

10-13 years14—18 years5 days2 months— 3 years4 — 6 years7 - 9 years

10-13 years14-18 years

81.3-233mg/l66.5-1 58 mg/155.1-1 13 mg/154.3 -130 mg/155.1 -113 mg/142.3- 99.1 mg/181.3-233 mg/166.5 -158 mg/155.1 -113 mg/154.3 -130 mg/155.1-113 mg/142.3- 99.1 mg/1

Free thyroxine

5 days2 months— 11 years12- 18 years

13.0-32.9 ng/l10.7-1 7.5 ng/l9.2 -15.9 ng/1

Eur. J. din·. Chem. Clin. Biochem. / Vol. 31,1993 / No. 5

288 Wiederaann et al.: Reference ranges for thyrotropin and thyroid hormones in paediatrics

An unequivocal age-dependent change in the serumconcentration of triiodothyronine is not evident fromthe collected results of other authors, although thereis a trend to lower values with increasing age. In thepresent study, serum triiodothyronine was signifi-cantly lower in neonates 5 days post partum than ininfants and children. A correlation analysis includingall age groups, however, showed a significant negativecorrelation between age and hormone concentration,i.e. the serum triodothyronine concentration de-creased with increasing age.

With the exception of Borkenstein et al. (8), all theother authors found age-related changes in serumthyroxine. An early marked decrease of serum thy-roxine in neonates is followed by a further downwardtrend in children and adolescents. In the present study,this downward trend was confirmed by a correlationanalysis between age and serum thyroxine concentra-tion.

The results of other authors consistently show thatserum concentrations of free thyroxine are markedlyhigher in neonates than in all other age groups. Nofurther apparent age-dependent changes were notedwith increasing age. In the present study, correlationanalysis showed a significant negative correlation be-tween age and free thyroxine, i.e. the serum concen-tration of free thyroxine decreased with increasingage. Statistically, this result is very probably influ-enced by the high concentrations of serum free thy-roxine in neonates 5 days post partum. No significantsex differences were found in any age group.

Table 13 shows the normal ranges for each hormone,as determined in the present study.

AcknowledgementWe thank Mrs. R. Biesenbach for technical assistance. AbbottLaboratories gave generous assistance, without which this workwould not have been possible.

References1. Deam, D., Goodwin, M. & Ratnaike, S. (1991) Comparison

of four methods for free thyroxin. Clin. Chem. 37, 569 —572.

2. Harkavy, K. L. & Enecio, C. E. (1991) Free thyroxine levelsin hospitalized newborns: Depressed levels in critical, non-thyroidal illness. J. Perinatol. XI, 117-121.

3. Kaptein, E. M., Macintyre, S. S., Weiner, J. M., Spencer,C. A. & Nicoloff, J. T. (1981) Free Thyroxine estimates innonthyroidal illness: Comparison of eighth methods. J.Clin. Endocrinol. Metab. 52, 1073-1077.

4. Midgley, J. E. M.? Sheehan, C. P., Christofides, N. D., Fry,J. E., Browning, D. & Mardell, R. (1990) Concentrationsof free thyroxin and albumin in serum in severe nonthy-roidal illness: Assay artefacts and physiological influences.Clin. Chem. 36, 765-771.

5. Rod, E., Gardini, E., Minelli, R., Bianconi, L. & Flisi, M.(1991) Thyroid function evaluation by different commer-cially available free thyroid hormone measurement kits interm pregnant women and their newborns. J. Endocrinol.Invest. 14, 1-9.

6. Witt, J. & Trendelenburg, C. (1982) Gemeinsame Studiezur Erstellung von Richtwerten klinisch-chemischer Kenn-größen im Kindesalter. J. Clin. Chem. Clin. Biochem. 20,235-242.

7. International Federation of Clinical Chemistry ScientificCommittee, Clinical Section Expert Panel on Theory ofReference Values and International Committee for Stan-dardisation in Haematology. Standing Comittee on Refer-ence Values (1987) Approved recommendation (1987) onthe theory of reference values. Part 5. Statistical treatmentof collected reference values. Determination of referencelimits. J. Clin. Chem. Clin. Biochem. 25, 645-656.

8. Borkenstein, M., Stöffler, G., Stogmann, W., Fueger, G. &Falk, W. (1980) Normwerte für zirkulierende Schilddrüsen-hormone, T3

iUptake und Thyrotropin vor und nach TRH-Gabe. Monatsschr. Kinderheilkd. 128, 422-427.

9. Struckmeyer, H. & Haid, H. (1986) Richtwerte für daskinderärztliche Laboratorium. 1. Auflage Marburg. DieMedizinische Verlagsgesellschaft.

10. Liappis, N. & Starke, A. (1987) Referenzwerte für dieKonzentration von freiem Thyroxin und freiem Trijodthy-ronin im Serum euthyreoter Kinder. Klin. Pädiatr. 199,366-369.

11. Liappis, N. & Schlebusch, H. (1988) Referenzwerte für dieKonzentration von Trijodthyronin, Thyroxin und Thyro-tropin im Blutserum euthyreoter Kinder. Methode: Lumi-neszenz-verstärkter Enzymimmunoassay. Klin. Pädiatr.200, 410-413.

12. Liappis, N., Schlebusch, H., Von Perjes, M. & Berg, I.(1991) Referenzwerte für die Konzentration des freien Thy-roxins, des freien Trijodthyronins und des Thyroxin-bin-denden Globulins im Blutserum euthyreoter Kinder. Me-thode: Lumineszenz-verstärkter Enzymimmunoassay. Klin.Pädiatr. 203, 113-115.

13. Schönberg, D. (1990) Hyopthyreose-Screening. Endokri-nologische Einzelschriften 3, 27-40.

14. Wiedemann, G. & Jonetz-Mentzel, L. (1992) Referenceranges for thyrotropin in the serum of full-term neonates— compared with the ranges for full-term neonates withvarious post-partal adaption disorders, and premature neo-nates. Eur. J, Clin. Chem. Clin. Biochem. 31, 35-39.

Doz. Dr. med. habil. Gudruri WiedemannKlinik und Poliklinik für Kindermedizinder Medizinischen Hochschule ErfurtAm Scheminbach 32 aO-5083 ErfurtBundesrepublik Deutschland


Establishment of Reference Ranges for Thyrotropin ...

Documents

Transcript of Establishment of Reference Ranges for Thyrotropin ...