Download Biologic Variation is Important for Interpretation of Thyroid Function

THYROID
Volume 13, Number 11, 2003
© Mary Ann Liebert, Inc.
Biologic Variation is Important for Interpretation of
Thyroid Function Tests
Stig Andersen, Niels Henrik Bruun, Klaus Michael Pedersen, and Peter Laurberg
Large variations exist in thyrotropin (TSH) and thyroid hormones in serum. The components of variation include preanalytical, analytical, and biologic variation. This is divided into between- and within-individual variation. The latter consists of circadian and seasonal differences although there are indicators of a genetically determined starting point. The ratio of within- to between-individual variation describes the reliability of
population-based reference ranges. This ratio is low for serum TSH, thyroxine (T4 ) and triiodothyronine (T3)
indicating that laboratory reference ranges are relatively insensitive to aberrations from normality in the individual. Solutions are considered but reducing the analytical variation below the calculated analytical goals of
7%, 5% and 12% for serum T3 , T4 , and TSH does not improve diagnostic performance. Neither does determination of the individual set-point and reference range. In practice this means that population-based reference
ranges are necessary but that it is important to recognize their limitations for use in individuals. Serum TSH
responds with amplification to minor alterations in T4 and T3 . A consistently abnormal TSH probably indicates
that T4 and T3 are not normal for the individual even when inside the laboratory reference range. This underlines the importance of TSH in diagnosis and monitoring of thyroid dysfunctions. Also, it implies that subclinical thyroid disease may be defined in purely biochemical terms. Under critical circumstances such as pregnancy where normal thyroid function is of importance for fetal brain development, subclinical thyroid disease
should be treated. Even TSH within the reference range may be associated with slightly abnormal thyroid function of the individual. The clinical importance of such small abnormalities in thyroid function in small children
and pregnant women for brain development remains to be elucidated.
Introduction
S
of hyperthyroidism and hypothyroidism are often nonspecific and vague if present (1–3)
and measurement of thyrotropin (TSH), thyroxine (T4), and
triiodothyronine (T3) in serum is important for diagnosis of
overt and subclinical thyroid dysfunction (4–8).
Usually, a test result in an individual is compared to a laboratory reference range. Such a reference range is often defined by a probability distribution including 95% of test results in healthy individuals in a population. Hence, the
variation in the population determines the width of the population-based reference range.
The overall variation consists of analytical and biologic
variation (9,10). Biologic variation is divided into two components: variation in the individual and variation between
individuals (9,10). The former is characterized by rhythmic
aberrations of multiple frequencies (11–14) and the latter is
caused by different set-points around which each individual
varies (9).
This paper aims to provide insight into the components of
YMPTO MS AND SIGNS
variation in TSH, T4, and T3 in serum, the determinants of
biologic variation, and the impact of biologic variation on interpretation of test results, specifically, the impact of variation on reference ranges and on alternative methods for detecting abnormal test results. Also, the largest acceptable
analytical variation in measurements of TSH, T4, and T3 in
serum is estimated. Finally, the impact on delineating the entity subclinical thyroid disease and the importance for individuals with TSH at the extremes of laboratory normal
ranges are estimated with consideration of the importance
for brain development.
Components of Variation in Thyroid Function Tests
Variation in thyroid function tests is described by a number of factors summarized in Table 1. A measured value, x,
is determined by an overall group mean, mgroup. Deviation
from this grand mean value is caused by the sum of components of variation. Minimizing each of these components
improves the precision of the measured value, x.
Department of Endocrinology and Medicine, University Hospital Aalborg, Aalborg, Denmark.
1069
1070
ANDERSEN ET AL.
TABLE 1. THE COMPONENT S INCLUDED IN A MEASUREMENT OF SERUM
TSH, T3 OR T4 CAN BE D ESCRIBED BY THIS EQUATION
x 5 mgroup 1 spreanalytical 1 sanalytical 1 sbetween-individual 1 swithin-individ ual 1 sother
x is a measured value of TSH, T3, or T4.
mgroup is the overall group mean value.
spreanalytical is variance due to preanalytical conditions.
sanalytical is variance due to analytical errors.
sbetween-individual is interindividual variance.
swithin-individ ual is intraindividual variance.
sother is random variation not accounted for.
Biologic variation consists of sbetween-individual 1 swithin-individual .
TSH, thyrotropin; T3 , triiodothyronine; T4, thyroxine.
Preanalytical variation, spreanalytical
Recommendations are that preanalytical conditions should
be standardized to minimize preanalytical variation in the
study of biologic variation (10,15): phlebotomy undertaken at
the same hour, by the same individual, without torniquet, in
a fasting, rested individual. Such conditions were applied in
two studies (16,17), inconsistently in one study (18), while information was limited in some studies (19–22).
Highly standardized preanalytical conditions may hamper
the external validity. Hence, one study of variation was performed to illustrate everyday testing of thyroid function with
blood samples taken during laboratory opening hours using
standard procedures (23). The larger variations found in this
study probably reflect the impact of preanalytical conditions.
ommended analytical coefficients of variation (CVs; 25–27).
The analytical variation is less important for TSH because
the biologic variation is quite high (27,28). The differences in
CVs reflect different preanalytical conditions.
Between-individual variation, sbetween-individual
Figure 1A shows mean values for each individual relative
to the overall mean value in a 1-year study of thyroid func-
A
Analytical variation, sanalytical
Analytical variation adds imprecision to biologic variation. Conversely, if biologic variation is high then analytical
variation becomes relatively less important and the requirement for analytical precision diminishes (10,24). The goals
set are that variance as a result of analytical error should not
exceed 20% of the variation (24). It can be shown that this is
acceptable when CVanalytical # 1/2 CVintraindividu al (10,24) or
# 1/2 CVbiologic (17).
Table 2 shows the analytical goals calculated from variation in a study using highly standardized preanalytical conditions (17), in a routine laboratory setting (23), and the rec-
B
TABLE 2. THE ANALYTICAL GOALS (CV%) CALCULATED FROM
DATA ON BIOLOGIC VARIATION IN A H IGHLY STANDARDIZED
SETTING (17), AND IN A ROUTINE LABORATORY SETTING
(CALCULATED FROM ANDERSEN 2002 [23]), AND THE
RECOMMEN DED MAXIMUM CV% (27)
Standardized
TT3
TT4
FT4I
TSH
a
Routine
Recommendations
CV% #a
CV% #a
CV% #b
CV% #b
5.2
2.5
4.7
8.1
6.8
5.1
5.8
11.6
10.2
9.7
10.2
21.9
7.9
7.0
7.7
28.6
Calculated from CVanalytical # 1/2 CVintraindivid ual (10,24).
Calculated from CVanalytical # 1/2 CVintraindividual 1 interindividual .
TT3 , total triiodothyronine; TT4 , total thyroxine; FT4 I, free
thyroxine index; TSH, thyrotropin.
b
FIG. 1. A: Individual mean in each of 15 healthy subjects
of estimated 24-hour urinary iodine excretion, thyrotropin
(TSH), total triiodothyronine (T3), and total thyroxine (T4).
Each dot represents the mean value of 12 consecutive
monthly measurements in 1 individual. Individual mean values are indexed around the group mean value (index 100:
urinary iodine 49.1 mg per 24 hours, serum TSH 1.27 mU/L,
T3 1.64 nmol/L, T4 106 nmol/L). (From Andersen et al.
[23,29]). B: Individual coefficients of variation (CV%) for each
variable. Each dot represents the CV% in one individual.
(From Andersen et al. [23,29]).
IMPACT OF VARIATION IN THYROID FUNCTION
1071
tion tests in 15 healthy men (23,29). This illustrates that each
individual had his own mean value around which the person varied. The difference between individual mean values
was highly significant (Kruskal-Wallis test; p , 0.001 for all
variables) (23). This is consistent with previous findings of
large variance between individuals (16–18,30).
Within-individual variation, swithin-individual
Repeated measurements of thyroid function tests in one
individual scatter around an individual mean value. This
variation differed widely between subjects (Fig. 1B, Bartlett’s
test, p , 0.001 for all variables) in our study (23) as found by
others (16–18).
The variation around the individual mean value is described by a probability distribution. The larger the difference between a measured value and the mean value, the less
likely is this caused by random variation. Thus, a 95% confidence interval in the individual and the distance required
between two measurements for significance can be calculated from data on variation.
What Determines Biological Variation in
Thyroid Function Tests
Circadian variation
A circadian variation in serum TSH in healthy subjects is
well documented (Table 3) (11–13, 31–50). Low levels of
TABLE 3. SOME STUDIES
OF
CIRCADIAN VARIATION
OF
serum TSH (nadir) are found during the daytime with a characteristic nocturnal rise of more than 100% that peaks just
after midnight. No increment of serum T3 or T4 seems to follow the nocturnal surge in TSH (32–35).
Lucke et al. (36) stated that “fluctuations do not exceed the
normal ranges.” This is the key point: these fluctuations contribute to the width of the reference ranges.
The circadian rhythm of TSH is influenced by environmental factors. Sleep (13,37,38) decreased TSH pulse amplitude but
not pulse frequency (13), as did fasting (39) and T3 and T4 infusion (40), while nocturnal physical activity may phase-delay
the rhythms (41). Interestingly, the circadian pulsatile secretion
pattern of serum TSH was found to be remarkably reproducible in the individual but with considerable and unaltered
differences between individuals (13,35). This favors a genetic
component in determining the pattern of TSH secretion.
Seasonal variation
Table 4 lists some longitudinal studies of seasonal variation in thyroid function tests in healthy adults (14,29,30,
51–67). In general, serum T3 was higher during winter while
seasonal changes in T4 and TSH were less consistent.
Seasonal variation has been suggested to be caused by ambient temperature and luminosity (51,52) and an increase in
T3 with cold has been demonstrated (68,69). Seasonal variation in iodine intake has been demonstrated in some populations (70–72) though not in all (73,74). Because iodine inTHYROID FUNCTION TESTS
IN
H EALTHY A DULT SUBJECTS
Increase from nadir to peak
Participants
(n)
7
26
8
5
6
10
5
7
24
5
10
6
5
8
10
8
5
13
6
Method for
data evaluation
Cosinor
Pulsec
Pulsec
Pulsec
Modified
Wilcoxon
Pulsec
—
ANOVA
Cosinor
—
Power-spectra
t test
F test, t test
Covariance
regression, t test
t test
ANOVA
—
t test
ANOVA
TSH
T3
T4
44%a
180%d
170%d
110%d
170%e
16%b
—
—
14%b
—
4%b
—
—
24%b
—
Persani et al., 1995 (33)
Brabant et al., 1990 (13)
Romijn et al., 1990 (37)
Brabant et al., 1987 (40)
Greenspan et al., 1986 (34)
150%d
320%d
250%d
87%d
120%d
170%a
97%f
14%d
96%d
—
—
—
—
22%b
35%a
20%b
22%d
—
—
—
—
—
—b
20%b
30%g
23%d
—
Brabant et al., 1986 (42)
Evans et al., 1986 (43)
Sowers et al., 1982 (44)
Custro and Scaglione, 1980 (45)
Weeke et al., 1980 (46)
Weeke and Gundersen, 1978 (12)
Chan et al., 1978 (47)
Lucke et al., 1977 (36)
Azukizawa et al., 1976 (35)
105%d
—
—
115%e
88%e
—
15%a
—
—
—
—
19%b
54%b
—
—
Parker et al., 1976 (38)
Balsam et al., 1975 (48)
O’Connor et al., 1974 (49)
Weeke, 1973 (11)
Patel et al., 1972 (50)
—, Data not included or specified.
a
p , 0.05.
b
p . 0.05.
c
Modified power spectrum analysis (42).
d
Significance level not specified or unclear.
e
p , 0.001.
f
p , 0.01.
g
p , 0.02.
TSH, thyrotropin; T3, triiodothyronine; T4, thyroxine; ANOVA, analysis of variance.
Reference
1072
ANDERSEN ET AL.
TABLE 4. SOME LONGITUDINAL STUDIES
OF
SEASONAL VARIATION
IN
THYROID FUNCTION TESTS
IN
HEALTHY ADULT SUBJECTS
Seasonal difference
Participants
(n)
16
20
13
43
166
13
6
32
4
15
45
53
5
13
8
12
9
8
18
4
Method for
data evaluation
TSH
T3
T4
Reference
ANOVA
ANOVA
Cosinor
ANOVA
Cosinor
Friedman’s
t test
Cosinor
ANOVA
t test
—
t test
Cosinor
Friedman’s
t test
—
ANOVA
—
—
—
8%a
31%b
25%c
—
21%c
16%a
—e
25%b
—
39%a
15%a
11%a
—
—
53%
—
—
—
—
—
5%a
13%b
6%c
77%c
18%b
—
17%b
18%e
12%b
7%d
5%a
27%c
—
9%d
22%a
14%a
4%a
—a
,1%a
35%c
3%a
—
22%e
4%a
—a
8%a
—
6%a
49%e
14%d
14%a
26%a
Andersen et al., 2001 (29)
Leppaluoto et al., 1998 (51)
Maes et al., 1997 (14)
Levine et al., 1995 (52)
Nicolau et al., 1992 (53)
Rasmussen et al., 1989 (54)
Harrop et al., 1985 (30)
Nicolau et al., 1984 (55)
Ponte et al., 1984 (56)
Konno 1980 (57)
Dessypris et al., 1981 (58)
Konno 1978 (59)
Reinberg et al., 1978 (60)
Smals et al., 1977 (61)
Rastogi and Sawhney, 1976 (62)
Postmes et al., 1973 (63)
Nuttall 1970 (64)
Ruisseau 1965 (65)
Thompson and Kignt 1963 (66)
Osiba 1957 (67)
18%a,f
48%f,g
18%f,g,h/87%f,g,i
68%f,g
—, Data not included or specified.
a
p . 0.05.
b
p , 0.05.
c
p , 0.001.
d
p , 0.02.
e
p , 0.01.
f
Protein-bound iodine in plasma.
g
Significance level not specified or unclear.
h
Working indoor (nurse students).
i
Working outdoor (postmen).
TSH, thyrotropin; T3, triiodothyronine; T4, thyroxine; ANOVA, analysis of variance.
fluences the secretion of thyroid hormones (75) some seasonal variation in thyroid function may be attributable to differences in iodine intake.
Statistically significant seasonal variations were not detected in our study (Fig. 2) although the trends found (29)
were similar to those statistically significant in studies with
a lower overall variation (14,30). Our study was performed
in an area with mild to moderate iodine deficiency and part
of the individual variation in serum TSH could be caused by
variation in iodine intake. Thus, we found a statistically significant trend in the average correlation between iodine excretion and TSH (Fig. 3) with a negative correlation between
urinary iodine and serum TSH in subjects with the lowest
average annual iodine intake (29). Probably this indicates a
variable lack of the substrate iodine for thyroid hormone synthesis in these subjects and thus different effects of variation
in iodine on thyroid function.
Other factors that may influence thyroid hormones and
TSH in serum
Gender differences was not found either in the pattern
(11–13,45,64) or in the variance (16–18) of TSH secretion. Protein concentrations in serum or their affinity to thyroid hormones may vary without changes in thyroid status (76). Prolonged venipressure and posture during phlebotomy (12)
may alter serum concentration of T3 and T4 slightly because
their concentrations relate to protein binding. This contrib-
utes to the difference in variance between studies (23) and
(16,17) but the effect is small. The level of TSH and thyroid
hormones in serum may be influenced by drugs (77,78)
smoking (79–81) and pregnancy (82,83) but little is known
about the effect on variation in thyroid function.
Impact of Biologic Variation on Interpretation of Tests
of Thyroid Function
Reliability of reference ranges
The ability of laboratory reference ranges to detect an abnormal test result in an individual depends on the balance
between contributors to the biologic variation. If the major
part of the overall variation is the result of variation within
individuals, while differences between individual set-points
are small, then a population-based reference range matches
variations in each of any individual. Conversely, if the overall variation is mainly caused by narrow individual variation around dispersed individual set-points then a population-based reference range is unlikely to detect minor
deviations from the individual set-point.
This can be described by a ratio of within- to between-individual variation, the individuality index (10,84,85). When
the ratio is less than 0.6, the population-based reference
range is an insensitive measure in the large majority of individuals. When the ratio is greater than 1.4, the reference
range works as intended (10,84).
IMPACT OF VARIATION IN THYROID FUNCTION
1073
FIG. 2. Twelve consecutive monthly estimated 24-hour urinary iodine excretion and measurements of serum thyrotropin
(TSH), total triiodothyronine (T3), and total thyroxine (T4) in 15 healthy subjects. Each measurement is divided by the individual mean value. (From Andersen et al. [23,29]).
Using highly standardized preanalytical conditions, this
ratio was below 0.6 for all tests of thyroid function (17). Using preanalytical conditions similar to those used in everyday laboratory practice the average individual variation was
only approximately half of that of the group (23) (Fig. 4). Accordingly, the ratio was still below 0.6 for all measures of
thyroid function (23). Consequently, an individual test result
may be far outside the individual reference range while still
lie well within the laboratory reference range. This indicates
a low sensitivity of the population-based reference ranges
and causes uncertainty in the diagnosis of overt, and in particular subclinical thyroid disease.
Individual reference ranges
FIG. 3. Average correlation coefficients (mean 6 standard
error [SE]) between estimated 24-hour urinary iodine excretion and serum thyrotropin (TSH) in 15 healthy men with
mild to moderate iodine deficiency. Participants were
grouped according to average annual estimated 24-hour urinary iodine excretion (nd 5 3, 6, 4, 2 in the four groups with
increasing iodine excretion). The average correlation decreased with increasing iodine excretion (p , 0.005 for
trend). Pearson correlation coefficients on ln-transformed
data were used. (From Andersen et al. [29]).
It has been suggested to use the individual as his own reference for detection of abnormal test results (18,86). To determine the individual set-point and reference range it is
necessary to analyze a number of specimens from that individual. The required number of measurements can be estimated if we decide on two things: the percentage closeness
to the true homeostatic set-point and the confidence interval. The number of tests required can then be calculated from
n 5 (Z 3 CVanalytical 1 intraindividu al/D)2, where n is the number of specimens, Z is 1.96 for the 95% confidence interval,
and D is the percentage closeness to the homoeostatic set
point (10).
Table 5 shows the average number of specimens required
to describe the homeostatic set-point in an individual in a
routine laboratory setting (23). The number of tests is large,
and because of the large differences in individual variance
(Fig. 1B) the number of tests required vary widely: from 30
to 183 tests of TSH, and from 2 to 126 measurements of T4
and T3 . Thus, this approach is both impractical and unreliable.
1074
ANDERSEN ET AL.
FIG. 4. The distribution of 12 consecutive monthly measurements of serum total thyroxine (T4) in 15 healthy subjects (n 5 180) and in 1 individual (n 5 12). The width of the
distribution in one individual is approximately half of that
of the group. (From Andersen et al. [23]).
Significant difference in repeated testing: the D check
A different way of evaluating a test result is to determine
whether this is significantly different from previous ones: the
delta check (D check). This may be calculated from D 5 Z 3
Ïn 3 ÏSDanalytical2 1 SDintraindividual 1 preanalytical2 , where n
is the number of specimens obtained and Z the confidence
interval (10,87).
A problem may occur if a mean intraindividual variance
is used for these calculations because of the large differences
in variance between individuals (p , 0.001 in all variables)
(23). Thus, an upper value was derived from the estimated
distribution of intraindividual variances rather than a mean
variance. Because only one type of deviation from mean variance was meaningful, a one-sided Z-value was used (0 for
50%, 1.64 for 95%, etc.). Hence, a true D check was computed,
and the differences required to be 50% to 99.9% confident of
an abnormal value compared to a previous test result in any
of the individuals can be read from Figure 5: As a rule of
thumb, a difference of approximately 40 nmol/L in T4, 0.7
nmol/L in T3, and 1.0 mU/L in TSH indicates 90% certainty
of a true change.
TABLE 5. NUMBER OF TESTS REQUIRED TO DESCRIBE
THE H OMEOSTATIC SET POINT IN AN INDIVIDUAL
(ADAPTED FROM A NDERSEN 2002 [23]).
Precision of set point
TSH
TT3
TT4
FT4I
5%
10%
20%
85
25
25
25
15
5
3
5
2
1
1
1
1
Calculated from: n 5 (Z*(CVanalytic al2 1 CVintraindividual 2 ) /2 /D)2
where:
D is % closeness to the homeostatic set point,
Z is the number of standard deviations required for a confidence
level (1.96 for 95%; 1.64 for 90%; 1,28 for 80%),
n is the number of specimens.
TSH, thyrotropin; TT3, triiodothyronine; TT4 , total thyroxine; FT4I,
free thyroxine index.
FIG. 5. The change in thyrotropin (TSH), total triiodothyronine (T3), total thyroxine (T4) and free T4 index required for
different levels of significance in repeated testing (the D check).
The degree of confidence of a statistically significant difference
from a previous measurement is read on the x-axis with the
corresponding difference required in TSH, TT3, TT4, and free
T4 index on the y-axis. This was calculated using D 5 Z 3
Ïn 3 ÏSDanalytical2 1 SDintraindividual 1 preanalytical 2, where n is the
number of specimens obtained and Z the two-sided confidence
interval (i.e., 0.67 for 50%, 1.64 for 90%, 1.96 for 95%). Derived
from Andersen et al. (23).
IMPACT OF VARIATION IN THYROID FUNCTION
This D check, or critical difference in serial testing, could
be perceived as a confidence interval around any individual
test result for each of the variables (compare Fig. 2).
Clinical Implications
Relation between TSH and thyroid hormones: a clue to
understanding subclinical thyroid disease
Serum TSH responds with amplification to minor changes
in serum T3 and T4 in our data (r 5 0.30 for log-linear relationship; r 5 0.17 for linear relationship) as described by others (88–91), and complete suppression of TSH could be found
when T3 was still within the normal range (88).
The position of the individual set point of T4 and T3 within
the wide laboratory reference ranges is important for the diagnosis of minor changes in thyroid hormones. If the set
point of an individual is in the upper part of the laboratory
reference range then only a slight increase in thyroid hormone secretion will cause these to leave the reference ranges,
as will the amplified response in TSH. This individual will
have diagnosed overt hyperthyroidism. If, however, the set
point of an individual is in the lower part of the laboratory
reference range (as in Fig. 4), then the same slight increase
in thyroid hormone secretion will not cause these to leave
the reference ranges while TSH will. This individual will
have diagnosed subclinical hyperthyroidism. Thus, narrow
individual variation combined with amplified response of
TSH to minor alterations in thyroid hormones in serum provides a biochemical explanation for the entity subclinical thyroid disease and indicates that this is mild thyroid failure.
Before week 12 of gestation the maturing brain is critically
dependent on the circulation maternal T4 and overt maternal hypothyroidism is associated with severely impaired
neurologic development of the offspring. More importantly,
even mild abnormalities in thyroid function may adversely
affect brain development in the fetus (92). Impaired intellectual development can be found even with T4 within the
reference range during early gestation (93). This emphasizes
the importance of detection of minor abnormalities in thyroid function.
Abnormalities with both thyroid hormones and TSH within
reference ranges
The ratio of within- to between-individual variation was
low for serum TSH in all studies (16–18,23). This implies that
some individuals with TSH within the population based reference range have a TSH outside the individual reference
range (i.e., an abnormal serum TSH).
This is illustrated in cross-sectional population studies in
which participants with serum TSH close to the outer limits
of the laboratory reference range have a higher frequency of
thyroid abnormalities. Thus, individuals with a high-normal
serum TSH have a higher occurrence of autoimmunity
(94,95) and a higher risk of developing thyroid disease (96).
Also, individuals with a low-normal serum TSH have more
nodules in the thyroid gland (97,98).
Thus, the individuals in the upper and lower parts of the
population-based reference ranges for TSH seem to consist of
two groups: those with normal thyroid function who just happen to have a high or a low set-point, and those who have
small abnormalities in thyroid function. Whether this influences brain development is an issue for future investigations.
1075
Impact of season and hour
Seasonal variation in serum TSH and thyroid hormones
can be found in Table 4. Changes in T3 vary with latitude
but on average serum total T3 increases approximately 0.2
nmol/L during winter. This is associated with an insignificant increase in serum TSH of approximately 0.1 mU/L.
Circadian variation in TSH is more pronounced (Table 3).
However, the main increase occurs in the evening and the
average decrease from 09.00 to 15.00 hours and may be estimated to be approximately 0.3 mU/L.
Conclusion and Recommendations
Individual variation in serum T4, T3, and TSH in healthy
subjects is narrow compared to laboratory reference ranges.
Consequently, a test result within the laboratory reference
range does not necessarily indicate a normal thyroid function in the individual. No mathematical trick may overcome
this problem because an impractically large number of tests
are required to determine the individual set-point. The distances between two measurements of thyroid function required for statistical significance can be seen in Figure 5 but
are quite large. However, serum TSH responds heavily to
minor changes in thyroid hormone concentrations in serum.
Hence, subclinical thyroid disease with abnormal TSH but
T4 and T3 within laboratory reference ranges is probably always a sign that T4 and T3 are outside the individual reference range and thus an indicatior for abnormal thyroid function in the individual. This emphasizes the importance of
serum TSH relative to T3 and T4 this being total or estimated
free hormone concentrations in serum. If there are clinical
signs, or if other conditions such as pregnancy requires normal thyroid function to ensure normal fetal brain development, then there is a need for treatment.
The individual reference range for serum TSH is approximately half the width of laboratory reference range. Thus,
some individuals in the upper and lower parts of the normal range will have TSH outside their individual reference
range. It remains to be determined whether such small abnormalities in thyroid function are important (i.e., for the developing brain).
References
1. Zulewski H, Müller B, Exer P, Miserez AR, Straub JJ 1997
Estimation of tissue hypothyroidism by a new clinical score:
Evaluation of patients with various grades of hypothyroidism and controls. J Clin Endocrinol Metab 82:771–776.
2. Smallridge RC 2000 Metabolic, physiologic, and clinical indexes of thyroid function. In: Braverman LE, Utiger RD (eds)
Werner & Ingbar’s The Thyroid: A Fundamental and Clinical Text. Lippincott Williams & Wilkins, Philadelphia, pp.
393–401.
3. Fatourechi V 2001 Subclinical thyroid disease. Mayo Clin
Proc 76:413–417.
4. Helfand M, Crapo L 1990 Monitoring therapy in patients
taking levothyroxine. Ann Intern Med 113:450–454.
5. Braverman LE 1996 Evaluation of thyroid status in patients
with thyrotoxicosis. Clin Chem 42:174–178.
6. Ridgway EC 1996 Modern concepts of primary thyroid
gland failure. Clin Chem 42:179–182.
7. Klee GG, Hay ID 1997 Biochemical testing of thyroid function. Endocrinol Metab Clin North Am 26:763–775.
8. Jarløv AE, Nygaard B, Hegedüs L, Hartling SG, Hansen JM
1076
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
1998 Observer variation in the clinical and laboratory evaluation of patients with thyroid dysfunction and goiter. Thyroid 8:393–398.
Harris EK, Kanofsky P, Shakarji G, Cotlove E 1970 Biological and analytic components of variation in long-term studies of serum constituents in normal subjects. II Estimating
biological components of variation. Clin Chem 16:1022–
1027.
Fraser CG, Harris EK 1989 Generation and application of
data on biological variation in clinical chemistry. Criti Rev
Clini Lab Sci 27:409–437.
Weeke J 1973 Circadian variations of the serum thyrotropin
level in normal subjects. Scand J Clin Lab Invest 31:337–342.
Weeke J, Gundersen HJG 1978 Circadian and 30 minutes
variations in serum TSH and thyroid hormones in normal
subjects. Acta Endocrinol 89:659–672.
Brabant G, Prank K, Ranft U, Schuermeyer T, Wagner TO,
Hauser H, Kummer B, Feistner H, Hesch RD, von zur
Mühlen A 1990 Physiological regulation of circadian and
pulsatile thyrotropin secretion in normal man and woman.
J Clin Endocrinol Metab 70:403–409.
Maes M, Mommen K, Hendrickx D, Peeters D, D’Hondt P,
Ranjan R, De Meyer F, Scharpé S 1997 Components of biological variation, including seasonality, in blood concentrations of TSH, TT3, FT4, PRL, cortisol and testosterone in
healthy volunteers. Clin Endocrinol 46:587–598.
Keffer JH 1996 Preanalytical considerations in testing thyroid function. Clin Chem 42:125–134.
Feldt-Rasmussen U, Petersen PH, Blaabjerg O, Hørder M
1980 Long–term variability in serum thyroglobulin and thyroid related hormones in healthy subjects. Acta Endocrinol
95:328–334.
Browning MCK, Ford RP, Callaghan SJ, Fraser CG 1986 Intra- and interindividual biological variation of five analytes
used in assessing thyroid function: Implications for necessary standards of performance and the interpretation of results. Clin Chem 32:962–966.
Nagayama I, Yamamoto K, Saito K, Kuzuya T, Saito T 1993
Subject based reference values in thyroid function tests.
Endo J 40:557–562.
Williams GZ, Widdowson M, Penton J 1978 Individual character of variation in time-series studies of healthy people II.
Differences in values for clinical chemical analytes in serum
among demographic groups, by age and sex. Clin Chem
24:313–320.
Watts GF, Pillay D, Kind PRN 1990 Influence of atenolol on
within-subject variation of thyroid function tests. Ann Clin
Biochem 27:599–600.
Juan-Pereira L, Navarro MA, Roca M, Fuentes-Arderiu X
1991 Within-subject variation of thyroxin and triiodothyronine concentrations in serum. Clin Chem 37:772–773.
Nicolau GY, Haus E 1992 Chronobiology of the hypothalamic-pituitary-thyroid axis. In: Touitou Y, Haus E (eds) Biologic Rhythms in Clinical and Laboratory Medicine.
Springer-Verlag, Berlin, pp. 330–347.
Andersen S, Pedersen KM, Bruun NH, Laurberg P 2002 Narrow individual variations in serum T4 and T3 in normal subjects: A clue to the understanding of subclinical thyroid disease. J Clin Endocrinol Metab 87:1068–1072.
Fraser CG 1983 Desirable performance standards for clinical chemistry tests. Adv Clin Chem 23:299–339.
Nelson JC, Wilcox RB 1996 Analytical performance of free
and total thyroxine assays. Clin Chem 42:146–154.
Klee GG 1996 Clinical usage recommendations and analytic
ANDERSEN ET AL.
performance goals for total and free triiodothyronine measurements. Clin Chem 42:155–159.
27. Demers LM, Spencer CA 2003 Laboratory medicine. Laboratory support for the diagnosis and monitoring of thyroid
disease. Thyroid 13:2–126.
28. Spencer CA, Takeuchi M, Kazarosyan M 1996 Current status and performance goals for serum thyrotropin (TSH) assays. Clin Chem 42:140–145.
29. Andersen S, Pedersen KM, Pedersen IB, Laurberg P 2001
Variations in urinary iodine excretion and thyroid function.
A 1-year study in healthy men. Euro J Endocrinol
144:461–465.
30. Harrop JS, Ashwell K, Hopton MR 1985 Circannual and
within-individual variation of thyroid function tests in normal subjects. Ann Clin Biochem 22:371–375.
31. Salvador J, Dieguez C, Scanlon F 1988 The circadian rhythms
of thyrotrophin and prolactin secretion. Chronobiol Int
5:85–93.
32. Fisher DA 1996 Physiological variations in thyroid hormones: Physiological and pathophysiological considerations. Clin Chem 42:135–139.
33. Persani L, Terzolo M, Asteria C, Orlandi F, Angeli A, BeckPeccoz P 1995 Circadian variations of thyrotropin bioactivity in normal subjects and patients with primary hypothyroidism. J Clin Endocrinol Metab 80:2722–2728.
34. Greenspan SL, Klibanski A, Schoenfeld D, Ridgway EC 1986
Pulsatile secretion of thyrotropin in man. J Clin Endocrinol
Metab 63:661–668.
35. Azukizawa M, Pekary AE, Hershman JM, Parker DC 1976
Plasma thyrotropin, thyroxine and triiodothyronine relationships in man. J Clin Endocrinol Metab 43:533–542.
36. Lucke C, Hehrmann R, von Mayersbach K, von zur Mühlen
A 1977 Studies on circadian variations of plasma TSH,
thyroxine and triiodothyronine in man. Acta Endocrinol
86:81–88.
37. Rossman LG, Parker DC, Pekary AE, Hershman JM 1979 Effect of an imposed 21 hour sleep-wake cycle upon the rhythmicity of human plasma thyrotropin. In: von Cauter E (ed)
Human Pituitary Hormones Circadian and Episodic Variations. A workshop symposium held in Brussels, Belgium,
November 29–30, pp. 96–113.
38. Parker DC, Pekary AE, Hershman JM 1976 Effect of normal
and reversed sleep-wake cycles upon noctohemeral rhythmicity of plasma thyrotropin: Evidence suggestive of an
inhibitory influence in sleep. J Clin Endocrinol Metab
43:318–329.
39. Romijn JA, Adriaanse R, Brabant G, Prank K, Endert E,
Wiersinge WM 1990 Pulsatile secretion of thyrotropin during fasting: A decrease of thyrotropin pulse amplitude. J Clin
Endocrinol Metab 70:1631–1636.
40. Brabant G, Brabant A, Ranft U, Ocran K, Köhrler J, Hesch
RD, von zur Mühlen A 1987 Circadian and pulsatile thyrotropin secretion in euthyroid man under the influence of
thyroid hormone and glucocorticoid administration. J Clin
Endocrinol Metab 65:83–88.
41. Buxton OM, Frank SA, L’Hermite-Balériaux M, Leproult R,
Turek FW, von Cauter E 1997 Roles of intensity and duration of nocturnal exercise in causing phase delays of human
circadian rhythms. Am J Physiol 273:E536–E542.
42. Brabant G, Ranft U, Ocran K, Hesch RD, von zur Mühlen A
1986 Thyrotropin: An episodically secreted hormone. Acta
Endocrinol 112:315–322.
43. Evans PJ, Weeks I, Jones MK, Woodhead JS, Scanlon MF
1986 The circadian variation of thyrotrophin in patients with
primary thyroidal disease. Clin Endocrinol 24:343–348.
IMPACT OF VARIATION IN THYROID FUNCTION
44. Sowers JR, Catania RA, Hershman JM 1982 Evidence for
dopaminergic control of circadian variations in thyrotropin
secretion. J Clin Endocrinol Metab 54:673–675.
45. Custro N, Scaglione R 1980 Circadian rhythm of TSH in
adult men and women. Acta Endocrinol 95:465–471.
46. Weeke J, Christensen SE, Hansen AP, Laurberg P, Lundbæk
K 1980 Somatostatin and the 24h levels of serum TSH, T3,
T4, and reverse T3 in normals, diabetics and patients treated
for myxoedema. Acta Endocrinol 94:30–37.
47. Chan V, Jones A, Liendo-Ch P, McNeilly A, Landon J, Besser
GM 1978 The relationship between circadian variations in
circulating thyrotrophin, thyroid hormones and prolactin.
Clin Endocrinol 9:337–349.
48. Balsam A, Dobbs CR, Leppo LE 1975 Circadian variations
in concentrations of plasma thyroxine and triiodothyronine
in man. J Appl Physiol 39:297–299.
49. O’Connor J, Wu GY, Gallagher TF, Hellman L 1974 The 24hour plasma thyroxin profile in normal man. J Clin Endocrinol Metab 39:765–771.
50. Patel YC, Alford FP, Burger HG 1972 The 24-hour plasma
thyrotropin profile. Clin Sci 43:71–77.
51. Leppaluoto J, Sikkila K, Hassi J 1998 Seasonal variation of
serum TSH and thyroid hormones in males living in subarctic environmental conditions. Int J Circumpolar Health
57:383–385.
52. Levine M, Duffy L, Moore DC, Matej LA 1995 Acclimation
of a non-indigenous sub-Arctic population: Seasonal variation in thyroid function in interior Alaska. Comp Bioch
Physiol 111:209–214.
53. Nicolau GY, Haus E, Plinga L, Dumitriu L, Lakatua D,
Popescu M, Ungureanu E, Sackett-Lundeen L, Petrescu E
1992 Chronobiology of pituitary-thyroid functions. Rom J
Endocrinol 30:125–148.
54. Rasmussen UF, Hegedüs L, Perrild H, Rasmussen N,
Hansen JM 1989 Relationship between serum thyroglobulin,
thyroid volume and serum TSH in healthy non-goitrous subjects and the relationship to seasonal variations in iodine intake. Thyroidology 3:115–118.
55. Nicolau GY, Lakatua D, Sackett-Lundeen L, Haus E 1984
Circadian and circannual rhythms of hormonal variables in
elderly men and women. Chronobiol Int 1:301–319.
56. Ponte AD, Guagnano MT, Sensi S 1984 Time-related behaviour of endocrine secretion: Circannual variations of FT3,
cortisol, HGH and serum basal insulin in healthy subjects.
Chronobiol Int 1:297–300.
57. Konno N 1980 Reciprocal changes in serum concentrations
of triiodothyronine and reverse triiodothyronine between
summer and winter in normal adult men. Endocrinol Jpn
27:471–476.
58. Dessypris A, Kirjarinta M, Miettinen A, Lamberg BA 1981
Seasonal variations of serum TSH, T3 and autoantibodies in
adult inhabitants of Lapland. Nordic Coun Arctic Med Res
Rep 33:603–605.
59. Konno N 1978 Comparison between the thyrotropin response to thyrotropin–releasing hormone in summer and
that in winter in normal subjects. Endocrinol Jpn 25:635–639.
60. Reinberg A, Lagoguey M, Cesselin F, Touitou Y, Legrand
JC, Delassalle A, Antreassian J, Lagoguey A 1978 Circadian
and circannual rhythms in plasma hormones and other variables of five healthy young human males. Acta Endocrinol
88:417–427.
61. Smals AGH, Ross HA, Kloppenborg PWC 1977 Seasonal
variation in serum T3 and T4 levels in man. J Clin Endocrinol
Metab 44:998–1001.
1077
62. Rastogi GK, Sawhney RC 1976 Thyroid function in changing weather in a subtropical region. Metabolism 25:903–908.
63. Postmes TJ, von Hout JC, Saat G, Willems P, Coenegracht J
1973 A radioimmunoassay study and comparison of seasonal variation in plasma triiodothyronine and thyroxine
concentrations in normal healthy persons. Clin Chim Acta
50:189–195.
64. Nuttall FQ 1970 Lack of seasonal variation in protein-bound
iodine. Clin Chim Acta 28:439–442.
65. De Ruisseau JP 1965 Seasonal variation of PBI in healthy
montrealers. J Clin Endocrinol Metab 25:1513–1515.
66. Thompson EM, Kight MA 1963 Effect of high environmental temperature on basal metabolism and concentrations of
serum protein-bound iodine and total cholesterol. Am J of
Clin Nutri 13:219–225.
67. Osiba S 1957 The seasonal variation of basal metabolism and
activity of thyroid gland in man. Jpn J Physiol 7:355–365.
68. Reed HL 1995 Circannual changes in thyroid hormone physiology: The role of cold environmental temperatures. Arctic
Med Res 54:9–15.
69. Hassi J, Sikkila K, Ruokonen A, Leppaluoto J 2001 The pituitary-thyroid axis in healthy men living under subarctic
climatological conditions. J Endocrinol 169:195–203.
70. Rasmussen LB, Ovesen L, Bülow I, Jørgensen T, Knudsen
N, Laurberg P, Perrild H 2002 Dietary iodine intake and urinary iodine excretion in a Danish population: Effect of geography, supplements and food choice. Br J Nutri 87:61–67.
71. Nelson M, Philips DI 1985 Seasonal variations in dietary iodine intake and thyrotoxicosis. Hum Nutr Appl Nutr
39A:213–216.
72. Nelson M, Phillips DIW, Morris JA, Wood TJ 1987 Urinary
iodine excretion correlates with milk iodine content in seven
British towns. J Epidem Comm Health 42:72–75.
73. Ford HC, Johnson LA, Feek CM, Newton JD 1991 Iodine intake and the seasonal incidence of thyrotoxicosis in New
Zealand. Clin Endocrinol 34:179–181.
74. Frey H, Rosenlund B, Storli U 1974 Urinary excretion of iodine in some norwegian population groups 1971–1972.
Tidsskrift for den Norske Lægeforening 30:982–987.
75. Pisarev MA, Gärtner R 2000 Autoregulatory actions of iodine. In: Braverman LE, Utiger RD (eds) Werner & Ingbar’s
The Thyroid: A Fundamental and Clinical Text. Lippincott
Williams & Wilkins, Philiadelphia, pp. 85–90.
76. Bartalena L, Robbins J 1992 Variations in thyroid hormone
transport proteins and their clinical implications. Thyroid
2:237–245.
77. Surks MI, Sievert R 1995 Drugs and thyroid function. N Engl
J Med 333:1688–1694.
78. Meier CA, Burger AG 2000 Effects of drugs and other substances on thyroid hormone synthesis and metabolism. In:
Braverman LE, Utiger RD (eds) Werner & Ingbar’s The Thyroid: A Fundamental and Clinical Text. Lippincott Williams
& Wilkins, Philiadelphia, pp. 265–280.
79. Vestergaard P 2002 Smoking and thyroid disorders: A metaanalysis. Euro J Endocrinol 146:153–161.
80. Brix TH, Hansen PS, Kyvik KO, Hegedüs L 2000 Cigarette
smoking and risk of clinically overt thyroid disease. Arch
Int Med 160:661–666.
81. Knudsen N, Bülow I, Laurberg P, Ovesen L, Perrild H, Jørgensen T 2002 Association of tobacco smoking with goiter
in a low-iodine-intake area. Arch Int Med 162:439–443.
82. Glinoer D 1999 What happens to the normal thyroid during
pregnancy. Thyroid 9:631–635.
83. Knudsen N, Laurberg P, Perrild H, Bulow I, Ovesen L, Jør-
1078
gensen T 2002 Risk factors for goiter and thyroid nodules. Thyroid 12:879–888.
84. Harris EK 1974 Effects of intra- and interindividual variation on the appropriate use of normal ranges. Clin Chem
20:1535–1542.
85. Harris EK 1981 Statistical aspects of reference values in clinical pathology. Prog Clin Pathol 8:45–66.
86. Winkel P 1977 Using the subject as his own referent in assessing day-to-day changes of laboratory test results. Contemp Top Anal Clin Chem 1:287–317.
87. Wheeler LA, Sheiner LB 1981 A clinical evaluation of various delta check methods. Clin Chem 27:5–9.
88. Pearce CJ, Himsworth RL 1986 Serum iodothyronine concentrations during introduction of thyroxine replacement
therapy in hypothyroidism. Clin Endocrinol 25:303–311.
89. Carr D, McLeod DT, Parry G, Thornes HM 1988 Fine adjustment of thyroxine replacement dosage: Comparison of
the thyrotrophin releasing hormone test using a sensitive
thyrotrophin assay with measurement of free thyroid hormones and clinical assessment. Clin Endocrinol 28:325–333.
90. Spencer CA, LoPresti JS, Patel A, Guttler RB, Eigen A, Shen
D 1990 Applications of a new chemiluminometric thyrotropin assay to subnormal measurement. J Clin Endocrinol
Metab 70:453–460.
91. Meier CA, Maisey MN, Lowry A, Müller J, Smith MA 1993
Interindividual differences in the pituitary-thyroid axis influences the interpretation of thyroid function tests. Clin Endocrinol 39:101–107.
92. Lazarus JH 1999 Thyroid hormone and intellectual development: A clinician’s view. Thyroid 9:659–660.
93. Haddow JE, Palomaki GE, Allan WC, Williams JR, Knight
GJ, Ganon J, O’Heir CE, Mitchell ML, Hermos RJ, Waisbren
SE, Faix JD, Klein RZ 1999 Maternal thyroid deficiency dur-
ANDERSEN ET AL.
94.
95.
96.
97.
98.
ing pregnancy and subsequent neuropsychological development of the child. N Engl J Med 341:549–555.
Hollowell JG, Staehling NW, Flanders WD, Hannon WH,
Gunter EW, Spencer CA, Braverman LE 2002 Serum TSH,
T4, and thyroid antibodies in the United States population
(1988 to 1994): national health and nutrition examination
survey (NHANES III). J Clin Endocrinol Metab 78:489–499.
Knudsen N, Jørgensen T, Rasmussen S, Christiansen E, Perrild H 1999 The prevalence of thyroid dysfunction in a population with borderline iodine deficiency. Clin Endocrinol
51:361–367.
Vanderpump MPJ, Tunbridge WMG, French JM, Appleton
D, Bates D, Clark F, Grimley EJ, Hasan D, Rogers H, Tunbridge F, Young ET 1995 The incidence of thyroid disorders
in the community: A twenty-year follow-up of the Whickham Survey. Clin Endocrinol 43:55–69.
Knudsen N, Bulow I, Jørgensen T, Laurberg P, Ovesen L,
Perrild H 2000 Comparative study of thyroid function and
types of thyroid dysfunction in two areas in Denmark with
slightly different iodine status. Euro J Endocrinol 143:485–491.
Knudsen N, Bulow I, Laurberg P, Perrild H, Ovesen L, Jørgensen T 2002 High occurrence of thyroid multinodularity
and low occurrence of subclinical hypothyroidism among
tobacco smokers in a large population study. J Endocrinol
175:571–576.
Address reprint requests to:
Stig Andersen, M.D., Ph.D.
Department of Endocrinology and Medicine
University Hospital Aalborg
Reberbansgade
Box 561
DK-9100 Aalborg
Denmark
E-mail: [email protected]
This article has been cited by:
1. Jesper Karmisholt , Stig Andersen , Peter Laurberg . 2008. Variation in Thyroid Function Tests in Patients with
Stable Untreated Subclinical Hypothyroidism. Thyroid 18:3, 303-308. [Abstract] [PDF] [PDF Plus]
2. Rolf Jorde , Oddmund Joakimsen , Eva Stensland , Ellisiv B. Mathiesen . 2008. Lack of Significant Association
between Intima-Media Thickness in the Carotid Artery and Serum TSH Level. The Tromsø Study. Thyroid
18:1, 21-25. [Abstract] [PDF] [PDF Plus]
3. Erdinc Serin, Kamil Gurel, Aysu Kiyan, Murat Caglayan, Erdal Karavas. 2008. Does diagnostic sonography alter
thyroid and parathyroid hormone levels?. Journal of Clinical Ultrasound 36:1, 12. [CrossRef]
4. Giampaolo Papi, Ettore degli Uberti, Corrado Betterle, Cesare Carani, Elizabeth N Pearce, Lewis E Braverman,
Elio Roti. 2007. Subclinical hypothyroidism. Current Opinion in Internal Medicine 6:4, 344-355. [CrossRef]
5. M. Klose, A. Juul, J. Struck, N. G. Morgenthaler, M. Kosteljanetz, U. Feldt-Rasmussen. 2007. Acute
and long-term pituitary insufficiency in traumatic brain injury: a prospective single-centre study. Clinical
Endocrinology, ahead of print070630051835004. [CrossRef]
6. Peter C. O'Leary, Peter H. Feddema, Valdo P. Michelangeli, Peter J. Leedman, Gerard T. Chew, Matthew
Knuiman, Joey Kaye, John P. Walsh. 2006. Investigations of thyroid hormones and antibodies based on a
community health survey: the Busselton thyroid study. Clinical Endocrinology 64:1, 97. [CrossRef]
7. Duk Jae Kim, Young Ho Khang, Jung-Min Koh, Young Kee Shong, Ghi Su Kim. 2006. Low normal TSH levels
are associated with low bone mineral density in healthy postmenopausal women. Clinical Endocrinology 64:1, 86.
[CrossRef]
8. Henry Völzke , Dietrich Alte , Thomas Kohlmann , Jan Lüdemann , Matthias Nauck , Ulrich John , Wieland
Meng . 2005. Reference Intervals of Serum Thyroid Function Tests in a Previously Iodine-Deficient Area. Thyroid
15:3, 279-285. [Abstract] [PDF] [PDF Plus]
9. Hock-Luen Teoh, Erle Chuen-Hian Lim. 2005. Platysmal myoclonus in subclinical hyperthyroidism. Movement
Disorders 20:8, 1064. [CrossRef]
10. Peter Laurberg, Stig Andersen, Inge B??low Pedersen, Allan Carl??. 2005. Hypothyroidism in the Elderly:
Pathophysiology, Diagnosis and Treatment. Drugs & Aging 22:1, 23. [CrossRef]
11. Terence T Lao. 2005. Thyroid disorders in pregnancy. Current Opinion in Obstetrics and Gynecology 17:2,
123???127. [CrossRef]
12. 2004. Letters to the Editor. Thyroid 14:11, 983-986. [Citation] [PDF] [PDF Plus]
13. Esther Jensen, Per Hyltoft Petersen, Ole Blaabjerg, Pia Skov Hansen, Thomas H. Brix, Kirsten Ohm Kyvik,
Laszlo Hegedüs. 2004. Establishment of a serum thyroid stimulating hormone (TSH) reference interval in healthy
adults. The importance of environmental factors, including thyroid antibodies. Clinical Chemistry and Laboratory
Medicine 42:7, 824-832. [CrossRef]