Download Large Differences in Incidences of Overt Hyper

0013-7227/02/$15.00/0
Printed in U.S.A.
The Journal of Clinical Endocrinology & Metabolism 87(10):4462– 4469
Copyright © 2002 by The Endocrine Society
doi: 10.1210/jc.2002-020750
Large Differences in Incidences of Overt Hyper- and
Hypothyroidism Associated with a Small Difference in
Iodine Intake: A Prospective Comparative RegisterBased Population Survey
INGE BÜLOW PEDERSEN, NILS KNUDSEN, TORBEN JØRGENSEN, HANS PERRILD,
LARS OVESEN, AND PETER LAURBERG
Department of Endocrinology and Medicine (I.B.P., P.L.), Aalborg Hospital, DK-9000 Aalborg, Denmark; Endocrine Unit,
Department of Internal Medicine I (N.K., H.P.), Bispebjerg University Hospital, DK-2400 Copenhagen, Denmark; Centre for
Preventive Medicine (N.K., T.J.), Glostrup Hospital, DK-2600 Glostrup, Denmark; and Institute of Food Research and
Nutrition (L.O.), the Danish Food Administration, DK-2860 Soeborg, Denmark
Around 3– 4 billion people in the world are covered by iodine
supplementation programs to prevent developmental brain
damage and other iodine deficiency (ID) disorders.
Mild ID is associated with more hyperthyroidism and less
hypothyroidism in the population than a high iodine intake.
Knowledge of the iodine intake levels where the shifts in incidences occur is important for planning of iodine supplementation programs.
A computer-based register linked to thyroid diagnostic laboratories was used to continuously identify all new cases of
overt hyper- and hypothyroidism in two population cohorts
with moderate and mild ID, respectively (Aalborg, n ⴝ 310,124;
urinary iodine, 45 ␮g/liter; and Copenhagen, n ⴝ 225,707; urinary iodine, 61 ␮g/liter). The investigation was initiated before iodization of salt in Denmark and was part of the monitoring program.
In 1997–1998, the incidence rate of overt hyperthyroidism
was high in the area with the lowest iodine intake (92.9/100,000
per year) compared with the area with only mild ID (65.4/
P
OPULATIONS WITH A LOW iodine intake from natural sources are found in all continents and most countries (1). When iodine deficiency (ID) is severe, thyroid hormone synthesis may be insufficient, and there is a high risk
of fetal and neonatal hypothyroidism with increased mortality, developmental brain damage, and a range of other
abnormalities (2, 3).
Iodine supplementation governed by national health care
agencies and guided by international organizations (International Council for the Control of Iodine Deficiency Disorders, World Health Organization, and United Nations Children’s Fund) has successfully reduced the risk of ID
disorders in most countries, with programs now covering on
the order of 3– 4 billion people in the world (4).
If the iodine supply to the thyroid is only moderately low,
the gland may compensate by a number of mechanisms to
maintain sufficient thyroid hormone production (5). Somehow, the compensating mechanisms may, over years, trigger
development of multifocal autonomous growth and function
Abbreviations: CI, 95% Confidence interval(s); FT4, free T4; GP, general practitioner; ID, iodine deficiency; SRR, standardized rate ratio; TT4,
total T4.
100,000 per year). Standardized rate ratio was 1.49, and 95%
confidence interval was 1.22–1.81. The opposite relationship
was present for overt hypothyroidism (moderate ID, 26.5/
100,000 per year; mild ID, 40.1/100,000 per year; standardized
rate ratio, 0.73; 95% confidence interval, 0.55– 0.97). The different incidence rates were confirmed during each of the two
following years.
The results of this prospective investigation of the incidence of overt hyper- and hypothyroidism suggest that iodine
supplementation of a population may increase the incidence
of overt hypothyroidism, even if the population is moderately
iodine-deficient. In such a population, the increase in risk of
hypothyroidism should be weighed against the risk of ID disorders such as hyperthyroidism due to multinodular toxic
goiter. The optimal level of iodine intake to prevent thyroid
disease may be a relatively narrow range around the recommended daily iodine intake of 150 ␮g. (J Clin Endocrinol
Metab 87: 4462– 4469, 2002)
of the thyroid gland with scattered cell clones harboring
activating mutations of the TSH receptor (6, 7). As a consequence, mild and moderate ID is associated with a high
incidence and prevalence of goiter and nodular hyperthyroidism in middle-aged and elderly subjects (8, 9).
A very high iodine intake may, however, also lead to
disease (10). Goiter and elevated serum TSH, indicating impaired thyroid function, are common in children and adults
with a high iodine intake (11–13). This may be due to inappropriate autoregulation of the thyroid by iodine. Excess
iodine inhibits the activity of several thyroidal processes,
possibly via synthesis of iodine containing arachidonic acid
derivatives (5). Moreover, iodine may enhance thyroid autoimmunity, leading to hypofunction of the gland (14 –17).
Hence, both low and high iodine intake of a population are
associated with an increase in the risk of thyroid disease,
albeit the consequences are more severe in ID than in iodine
excess (10, 18).
It is important for planning of public iodine supplementation programs to know at which level of iodine intake the
shift from a high incidence of nodular hyperthyroidism to a
high incidence of thyroid hypofunction takes place.
4462
12.6– 40.4
41.7– 83.9
52.5–101.3
100.5–167.1
144.3–239.9
197.7–322.3
316.1–532.7
83.0–102.8
0.2– 8.8
3.3–16.7
0–5.6
9.7–34.7
0–28.5
16.0– 48.8
25.3–73.9
68.4–148.8
49.9–153.5
21.2–31.8
2.2
4.9
1.9
22.2
14.2
32.4
49.6
108.6
100.7
26.5
Hypo
Hypo
Hyper
95% CId
0
0
26.5
62.8
76.9
133.8
192.1
260.0
424.4
92.9
a
0
0
3.6
21.5
31.9
38.7
96.6
132.9
225.5
36.0
0
0
0
1
0
2
1
2
3
9
0
1
3
3
12
15
26
24
13
97
1
2
1
13
7
17
17
30
17
105
39,321 (48.7/51.3)
35,311 (48.7/51.3)
45,362 (47.7/52.3)
46,375 (48.4/51.6)
42,375 (49.3/50.7)
39,715 (49.8/50.2)
27,659 (51.9/48.1)
22,090 (56.2/43.8)
11,916 (64.9/35.1)
310,124 (50.2/49.8)
0 –9
10 –19
20 –29
30 –39
40 – 49
50 –59
60 – 69
70 –79
⬎80
All ages
0
1
17
37
50
77
88
91
72
433
Hyper
Hypo
Hyper
Hypo
Hyper
Results from the 14 months of observation in Aalborg (March 1, 1997, to April 30, 1998). F, Female; M, male.
Number of cases from database and contact to GP.
b
Number of cases excluded during final evaluation.
c
Incidence rate: number of new cases/100,000 per year.
d
95% CI of the total incidences of hyper- and hypothyroidism.
4.5
10.0
4.0
42.0
28.7
56.4
71.7
145.0
121.9
43.5
0
0
51.5
107.0
123.1
229.9
280.8
359.0
532.0
149.1
0
0
0
3.6
0
8.6
25.8
62.0
61.5
9.4
Hypo
Hyper
Hypo
Hyper
Total
Female
Male
Study population
Total (F/M %)
Details of the register and the methodological evaluations performed
have been published recently (25). In brief, all blood tests from subjects
living in the two study areas were analyzed in one of four clinical
chemistry laboratories (one in Aalborg, three in Copenhagen). Results
of TSH and thyroid hormone measurements were continuously imported from the laboratory databases into a register database, which
automatically identified overt hyperthyroidism (low serum TSH combined with a high serum T3 and/or serum T4), and overt hypothyroidism
(high serum TSH combined with a low serum T4). A list of not previously
registered cases of biochemical hyper- and hypothyroidism was generated each day and evaluated by: 1) searching in laboratory databases
and diagnosis registers of the hospitals in the area, 2) rechecking that the
patient lived in one of the two geographical areas, and 3) contacting the
patients’ general practitioner (GP) for verification. All new cases were
subsequently invited to an interview and physical examination with
blood sampling for a final evaluation.
Before the registration was initiated, various details of the methodology were evaluated (25). This included evaluation of the diagnostic
performance of the participating laboratories. A reference panel of normal sera (n ⫽ 100) was analyzed in the four laboratories. No systematic
differences in distribution of values with reference ranges were found.
The database identification of new cases was tested by a 60-d run in
Age (yr)
Identification of new cases of hyper- and hypothyroidism
Incidence ratec
The present study is part of The Danish Investigation of Iodine Intake
and Thyroid Diseases (DanThyr), which comprises a number of register
studies and a cohort study. DanThyr is the official clinical monitoring
of the Danish iodine supplementation program, and it is projected to run
for 9 yr in its present form. The present register study measured prospectively the incidence rates of hyper- and hypothyroidism in two
population cohorts with mild and moderate ID before iodine
supplementation.
Two open population cohorts representing the geographical variation
in iodine intake in Denmark (19) were selected for monitoring. One
population cohort consisted of 310,124 subjects living in the city of
Aalborg and the surrounding municipalities in Northern Jutland. This
area had moderate ID with a median iodine concentration in urine of 45
␮g/liter (estimated iodine excretion, 62 ␮g/24 h) in subjects taking no
supplements. The other population cohort comprised 225,707 subjects
living in the geographical area around Bispebjerg and Frederiksberg
Hospital in Copenhagen. The area had mild ID with a median urinary
iodine concentration of 61 ␮g/liter (estimated iodine excretion, 93 ␮g/24
h) when no supplements were taken. Median urinary iodine values were
moderately higher if all subjects were included (Aalborg, 53 ␮g/liter;
Copenhagen, 68 ␮g/liter). The iodine excretion values were obtained in
a simultaneous cohort study of 4649 subjects from the two areas (21, 22).
The participants had been randomly selected in certain age and sex
groups from the general population (females aged 18 –22, 25–30, 40 – 45,
and 60 – 65 yr; and males aged 60 – 65 yr). In all groups, the urinary iodine
excretion was higher in Copenhagen than in Aalborg (21).
Age and sex compositions of the present population cohorts at January 1, 1998, are given in Tables 1 and 2. Information on the size and
composition of population cohorts on January 1, 1998, 1999, and 2000
was obtained from the Danish Bureau of Statistics. The Danish population is relatively homogeneous, and the population iodine intake has
been rather stable over at least 30 – 40 yr, with the exception of a subgroup now taking vitamin mineral tablets containing iodine (23, 24).
Excluded during
verificationb
Subjects and Methods
Population cohorts
New cases from
databasea
In this prospective study, we compared the long-term
effects of different grades of low iodine intake on the incidence of overt thyroid function abnormalities in two population cohorts with different iodine intake levels (moderate
and mild ID) (19) due to differences in water iodine contents
in the two localities (20). The results suggest that any increase
in iodine intake of a population that is not severely iodine
deficient may induce an increase in the incidence of
hypothyroidism.
J Clin Endocrinol Metab, October 2002, 87(10):4462– 4469 4463
TABLE 1. Total number and incidence rates of hyper- and hypothyroidism according to age and sex groups in an open population cohort from Aalborg with moderate ID
Pedersen et al. • Thyroid Disease and Iodine Intake
1.2–17.4
0.2–18.4
4.3–39.1
17.6– 68.4
57.9–152.5
44.6–125.6
152.5–303.5
32.2– 48.0
8.3–32.5
22.5–19.3
52.1–121.7
31.9–93.3
79.7–186.1
117.1–233.3
136.2–280.6
55.3–75.5
0–12.3
3.9
0
9.3
9.3
21.7
43.0
105.2
85.1
228.0
40.1
Evaluation of cases
a
0
0
3
2
2
1
5
8
4
25
1
0
1
7
5
3
4
11
9
41
1
0
12
25
29
19
28
46
41
201
0 –9
10 –19
20 –29
30 –39
40 – 49
50 –59
60 – 69
70 –79
⬎80
All ages
23,447 (48.7/51.3)
14,397 (49.7/50.3)
49,756 (52.4/47.6)
39,721 (46.5/53.5)
25,502 (48.3/51.7)
23,605 (50.2/49.8)
16,674 (55.5/44.5)
18,433 (64.0/36.0)
14,172 (74.1/25.9)
225,707 (52.7/47.3)
1
0
8
6
8
12
24
25
39
123
Hypo
Hyper
Hypo
Hyper
Study population
Total (F/M %)
Age (yr)
Results from the 13 months of observation in Copenhagen (May 1, 1997, to May 31, 1998). F, Female; M, male.
Number of cases from database and contact to GP.
b
Number of cases excluded during final evaluation.
c
Incidence rate: number of new cases/100,000 per year.
d
95% CI of the total incidences of hyper- and hypothyroidism.
0
0
20.4
41.8
86.9
62.6
132.9
175.2
208.4
65.4
0
0
17.7
15.0
29.9
85.7
169.5
86.1
237.4
60.6
7.7
0
0
4.3
14.0
0
24.9
83.4
201.1
17.3
0
0
3.9
13.0
56.1
15.7
49.8
152.9
50.3
26.8
0
0
35.4
74.9
119.8
109.0
199.4
187.9
263.8
101.7
Hyper
Hypo
Hyper
Hypo
Hyper
Female
Male
Incidence ratec
Excluded during
verificationb
New cases from
databasea
Pedersen et al. • Thyroid Disease and Iodine Intake
Aalborg with parallel manual and computer-based evaluation of all
abnormal TSH results (n ⫽ 2159). The results obtained were identical.
Furthermore, we evaluated the risk to be lost from detection because of
GPs outside the study area. This was found to be negligible. Finally, the
use of TSH as the primary test in the diagnostic algorithm was evaluated.
No cases of thyroid dysfunction had been diagnosed based on T3 and
T4 alone.
Total
Hypo
Hyper
95% CId
Hypo
J Clin Endocrinol Metab, October 2002, 87(10):4462– 4469
TABLE 2. Total number and incidence rates of hyper- and hypothyroidism according to age and sex groups in an open population cohort from Copenhagen with mild ID
4464
In two of the laboratories in Copenhagen, serum free T4 (FT4) was
analyzed in addition to serum total T4 (TT4). The other laboratories relied
on supplementary measurements of serum thyroid binding globulin or
a T3 test to evaluate abnormalities of protein binding of T4 and T3 in
serum. Because this was a comparative study, cases registered on condition of an abnormal FT4 combined with a normal TT4 were excluded
to avoid bias [cases with low TSH combined with normal TT3 and TT4
but high FT4 (n ⫽ 7) or high TSH combined with normal TT4 but low
FT4 (n ⫽ 16)]. Cases with abnormal FT4 and missing TT4 (n ⫽ 6) were
included.
During the final evaluation, another 149 apparently new cases were
excluded for the following reasons: normalization of thyroid hormones
(serum T4 and T3) within 3 wk without treatment or clinical signs of acute
or subacute thyroiditis (n ⫽ 105); abnormal protein binding of thyroid
hormones (n ⫽ 3); amiodarone treatment with low TSH and high TT4
but normal TT3 (n ⫽ 16); cases primarily verified by GP to be new,
despite earlier treatment for thyroid disease (n ⫽ 14); treatment with
levo-T4 for a psychiatric disorder (n ⫽ 4); substitution with levo-T4 after
thyroid surgery (n ⫽ 3); overtreatment of subclinical hyperthyroidism
with carbimazole (n ⫽ 1); and laboratory errors (n ⫽ 3). Such detailed
evaluation with new blood tests and clinical follow-up was performed
in 94.4% of the patients in Aalborg and 88.5% in Copenhagen. Information on cases excluded during evaluation is given in Tables 1 and 2.
From the persons evaluated in this way, we calculated for each
population cohort the additional fraction and number of patients with
biochemical hyper- and hypothyroidism that hypothetically would have
been excluded in each sex and age group, if follow-up had been performed in all patients. Adjustment of incidence values using these data
did not alter the values substantially, and no difference in statistical
results occurred.
Registration after introduction of iodized salt
The incidence rates of hyper- and hypothyroidism in the study period
and during the two subsequent 1-yr periods after introduction of iodine
supplementation were compared. The incidence rates were calculated
from the number of new cases identified by the register database and
verified by contact to GPs, but without follow-up evaluation of patients.
Laboratory activity
The register accumulated information on the number of thyroid laboratory tests in the two population cohorts performed as part of diagnosis and control of thyroid disorders. Overall, the activity was similar
in the two areas and moderately increasing during the time period
studied. The number of TSH analyses was slightly higher in Copenhagen, whereas more T4 and T3 analyses were performed in Aalborg
(Table 3).
TSH, T3 and T4 analyses used
Aalborg Hospital Laboratory. TSH, immunoluminometric assay (Lumitest,
Brahms Diagnostica, Berlin, Germany); reference interval, 0.3– 4.5 mU/
liter. T3 and T4, competitive RIA (Amerlex-m T3 and T4 RIA Kit, OrthoClinical Diagnostics, Raritan, NJ); reference T3, 1.2–2.7 nmol/liter; T4,
60 –140 nmol/liter.
Frederiksberg Hospital Laboratory. TSH, time-resolved fluoroimmunometric assay based on direct sandwich technique (AutoDelfia hTSH Ultra
kit, Wallac, Inc., Gaithersburg, MD); reference, 0.15– 4.5 mU/liter. T3 and
T4, time-resolved fluoroimmunoassay based on competitive reaction
(AutoDelfia T3 and T4 kit, Wallac, Inc.); reference T3, 0.9 –2.7 nmol/liter;
T4, 60 –160 nmol/liter.
Pedersen et al. • Thyroid Disease and Iodine Intake
J Clin Endocrinol Metab, October 2002, 87(10):4462– 4469 4465
TABLE 3. Thyroid laboratory tests per 100,000 inhabitants per year in the open cohorts from Aalborg (moderate ID) and Copenhagen
(mild ID) before and the first 2 yr after iodized salt was introduced
Aalborg
Serum
Serum
Serum
Serum
TSH
T3
T4
FT4b
Copenhagen
Before
1st yr aftera
2nd yr aftera
Before
1st yr aftera
2nd yr aftera
19,368
9,654
11,340
0
21,245
11,072
11,618
0
23,051
12,502
13,006
0
22,345
7,576
7,630
2,714
23,624
7,538
7,578
2,823
24,408
7,500
7,548
3,012
Number of tests per 100,000 inhabitants per year performed by the laboratories in the two cohort areas.
Period in relation to iodization of salt. Before: Aalborg, March 1, 1997, to April 30, 1998; Copenhagen, May 1, 1997, to May 31, 1998. 1st
yr after: June 1, 1998, to May 31, 1999. 2nd yr after: June 1, 1999, to May 31, 2000.
b
Many Danish laboratories are not using measurements of FT4 and FT3 in serum for routine purposes due to the insufficient methodology
of most commercial methods. In this study, FT4 was only measured in two of the laboratories in Copenhagen.
a
FIG. 1. The incidence rates of hyper- and hypothyroidism in males and females in two open population cohorts with different levels of iodine
intake (Aalborg (f) with moderate ID, and Copenhagen (䡺) with mild ID). The incidence rates were calculated after follow-up evaluation of
patients. Both hyper- and hypothyroidism (age standardized) were more common in females than in males in both cohorts [hyperthyroidism,
Aalborg, SRR (females/males), 3.24; CI, 2.36 – 4.44; Copenhagen, SRR, 3.02; CI, 1.99 – 4.56; and hypothyroidism, Aalborg, SRR (females/males),
3.18; CI, 1.75–5.78; Copenhagen, SRR, 1.97; CI, 1.17–3.30]. The incidence rate of hyperthyroidism was higher in Aalborg than in Copenhagen
[SRR (Aalborg/Copenhagen) total, 1.49; CI, 1.22–1.81; females, SRR, 1.65; CI, 1.34 –2.04; males, SRR, 1.54; CI, 0.96 –2.46]. The incidence rate
of hypothyroidism was higher in Copenhagen than in Aalborg [SRR (Aalborg/Copenhagen) total, 0.73; CI, 0.55– 0.97; females, SRR, 0.79; CI,
0.57–1.09; males, SRR, 0.49; CI, 0.24 –1.0].
Bispebjerg Hospital Laboratory. TSH, microparticle enzyme immunoassay
(Axsym ultrasensitive hTSH II, Abbott Laboratories, Abbott Park, IL);
reference, 0.15–5.0 mU/liter. T3, microparticle enzyme immunoassay
(Axsym T3, Abbott Laboratories); reference, 1.2–2.3 nmol/liter. T4, fluorescence polarization immunoassay (Axsym, Abbott Laboratories); reference, 60 –140 nmol/liter. FT4, microparticle enzyme immunoassay
(Axsym FT4, Abbott Laboratories); reference, 9.0 –24.0 pmol/liter.
The Laboratory of General Practitioners in Copenhagen. TSH, two-sided
chemiluminometric immunoassay (Ciba Corning ACS TSH, Ciba Corning Diagnostic, Medfield, MA); reference, 0.2–5.0 mU/liter. T3, T4, and
FT4, competetive immunoassay (Ciba Corning ACS T3, ACS T4, and ACS
FT4, all by Ciba Corning Diagnostic); reference T3, 1.1–2.6 nmol/liter; T4,
60 –140 nmol/liter; and FT4, 10.0 –22.0 pmol/liter.
All involved TSH assays had a functional sensitivity below 0.07
mU/liter.
Statistical methods
The 95% confidence intervals (CI) for standardized rate ratios (SRRs)
were calculated after log transformation of the respective rates (26).
When comparing the total incidence rates, the analyses were performed
after age and sex standardizing of the Aalborg cohort to the Copenhagen
cohort. Female and male incidence rates were compared after age standardizing to the Copenhagen females (Fig. 1). Standardization to a
female-to-male ratio of 1:1 was performed when comparing incidence
rates between different age groups (Fig. 2). The calculations were based
on an assumption of Poisson distribution of cases. Level of significance
was set to 5%.
The cumulative risk was calculated by a summation of the agespecific incidence rates multiplied by 10, because the risk for each year
within the 10-yr age group was assumed to be constant. No correction
for prevalent cases was attempted.
4466
J Clin Endocrinol Metab, October 2002, 87(10):4462– 4469
Pedersen et al. • Thyroid Disease and Iodine Intake
FIG. 2. The incidence rates of hyper- and hypothyroidism in different age groups in Aalborg (f) with moderate ID and Copenhagen (䡺) with
mild ID. In both cohorts, the sex standardized incidence rates of hyper- and hypothyroidism increased (data not shown). Differences in incidence
rates between the Aalborg and Copenhagen cohorts were statistically significant in the age group above 60 yr [hyperthyroidism, SRR
(Aalborg/Copenhagen), 1.62; CI, 1.26 –2.09; and hypothyroidism, SRR, 0.63; CI, 0.44 – 0.89]. In the other age groups, only nonsignificant
tendencies were observed (data not shown).
The study was approved by the regional Ethics Committees in Northern Jutland and Copenhagen.
Results
The primary investigation took place in 1997–98 and lasted
14 months in Aalborg and 13 months in Copenhagen, until
iodized salt was introduced in June 1998. Subsequently, data
from the first 2 yr after iodization of salt were included.
Details on the number of new patients are given in Tables 1
and 2.
The crude incidence rate of overt hyperthyroidism (with
no correction for age and sex distribution) was high in Aalborg (92.9/100,000 per year) compared with Copenhagen
(65.4/100,000 per year). On the other hand, overt hypothyroidism was more common in Copenhagen (40.1/100,000 per
year) than in Aalborg (26.5/100,000 per year). There were
small differences in the age composition of the two population cohorts (Tables 1 and 2). Age and sex standardization
of the Aalborg cohort to the Copenhagen cohort affected the
incidence rates little, and differences between cohorts were
statistically significant (hyperthyroidism, SRR (Aalborg/
Copenhagen), 1.49; CI, 1.22–1.81; hypothyroidism, SRR, 0.73;
CI, 0.55– 0.97).
In both areas, hyper- and hypothyroidism were more common in females than in males (Fig. 1), and the incidence rates
of both hyper- and hypothyroidism increased with age (Fig.
2). The pattern of difference between the population cohorts
TABLE 4. Cumulative risk (%) of having diagnosed thyroid
dysfunction before age 90 yr in Aalborg with moderate ID and
Copenhagen with mild ID
Hyperthyroidism
Hypothyroidism
Hyper- or hypothyroidism
Aalborg (F/M)
Copenhagen (F/M)
11.8 (16.8/5.5)
3.4 (4.8/1.6)
15.2 (21.6/7.1)
7.3 (9.9/3.4)
5.1 (6.4/3.4)
12.4 (16.3/6.8)
Cumulative risk was calculated up to 90 yr of age from the incidence rates given in Tables 1 and 2. F, Female; M, male.
was the same in nearly all subgroups; hyperthyroidism was
more frequent in the cohort with moderate ID, whereas hypothyroidism was more frequent in the cohort with mild ID.
The cumulative risk of having diagnosed thyroid dysfunction up to the age of 90 yr is shown in Table 4. The highest
risk (21.7%) was found in females from Aalborg.
Figure 3 depicts the incidence rates of hyper- and hypothyroidism in the primary study period and during the two
subsequent 1-yr periods. The incidence rates given were
calculated from the number of new cases identified by the
register database and verified by contact to GPs, but without
follow-up evaluation of patients. The differences in incidence
rates of hyper- and hypothyroidism between population cohorts persisted in the 2 yr of follow-up. Over the entire study
period, the incidence rate of thyroid dysfunction (hyper- plus
hypothyroidism) was 163.8/100,000 per year in Aalborg and
Pedersen et al. • Thyroid Disease and Iodine Intake
J Clin Endocrinol Metab, October 2002, 87(10):4462– 4469 4467
FIG. 3. The incidence rates of hyper- and hypothyroidism in the two open population cohorts with moderate ID (Aalborg) and mild ID
(Copenhagen) before and after iodized salt was introduced in Denmark by June 1998. The rates were calculated from the number of new cases
identified by the register database and verified by contact with GPs, but without follow-up evaluation of patients. The differences in incidence
rates of hyper- and hypothyroidism between the two cohorts were significant in all three periods. Hyperthyroidism: before salt iodization, SRR
(Aalborg/Copenhagen), 1.62; CI, 1.37–1.93; first year after, SRR, 1.79; CI, 1.52–2.12; second year after, SRR, 1.94, CI, 1.64 –2.30. Hypothyroidism: before salt iodization, SRR (Aalborg/Copenhagen), 0.66; CI, 0.51– 0.87; first year after, SRR, 0.65; CI, 0.49 – 0.85; second year after,
SRR, 0.65; CI, 0.49 – 0.85. During the 2 yr after salt iodization, the use of iodized salt and the increase in iodine intake was low (on average,
⬍5 ␮g/d). Compulsory iodization has now been introduced.
133.4/100,000 per year in Copenhagen (SRR, 1.35; CI,
1.25–1.47).
In the Aalborg population cohort, an increase in incidence
rate of hyperthyroidism was seen in the second year after
iodine supplementation (SSR before vs. second year after
supplementation, 0.74; CI, 0.68 – 0.81). No significant alterations were seen in incidence rates of hyperthyroidism in
Copenhagen or in hypothyroidism in either population
cohort.
Discussion
This prospective comparative epidemiological study
showed that the incidence rate of overt hyperthyroidism was
much higher in a population cohort with moderate ID than
in an otherwise comparable cohort with mild ID. The opposite relationship was present for overt hypothyroidism,
where the highest incidence rate was found in the area with
mild ID.
Before the recent implementation of widespread iodine
supplementation, ID was the most common cause of preventable brain damage in the world (2, 3), and public iodine
supplementation programs are thus important parts of
global health care. In areas less severely affected, such as
Denmark, a low iodine intake predominantly manifests with
a high risk of goiter and autonomous thyroid nodules with
hyperthyroidism (27). The primary purpose of iodine supplementation in such areas is to prevent goiter and nodular
hyperthyroidism. The present study indicates that considerable prevention can be expected already at a relatively low
level of iodine supplementation. This is supported by our
previous findings in a comparative cohort study (n ⫽ 4649)
from the same two areas with moderate and mild ID. The
prevalence of goiter was considerably higher in the area with
moderate ID (22). Moreover, in moderate ID we observed a
gradual decrease in median serum TSH with age, as a sign
of development of autonomous thyroid nodules. In mild ID,
this was much less prominent (28).
The effect of iodine on the thyroid gland is complex, with
a U-shaped relation between iodine intake and risk of thyroid
disease, because both low and high iodine intake are associated with an increased risk (3, 10, 18). When iodine intake
is high, the prevalence of hypothyroidism and diffuse thyroid enlargement is high compared with populations with
iodine intake around the recommended level of 150 ␮g/d
(11). The mechanism behind this impairment of thyroid function is unknown, but both iodine enhancement of thyroid
autoimmunity and reversible iodine inhibition of thyroid
function in susceptible subjects are possibly involved (5,
14 –17).
There is only limited knowledge of the threshold of iodine
intake, above which the increase in hypothyroidism begins.
In the present study, we found a high incidence rate of
hypothyroidism in Copenhagen with mild ID compared
with Aalborg with moderate ID. This suggests that the increase in hypothyroidism, induced by an increase in iodine
intake, may start already at low levels of iodine intake. The
increase in hypothyroidism may continue over a wide range
of iodine intake levels because it was observed in a Chinese
study that the most prominent consequence of a change in
iodine intake from normal (⬃150 ␮g/d) to moderately high
(⬃500 ␮g/d) was a severalfold increase in the prevalence of
thyroid hypofunction (29, 30).
In our study, new cases of hyper- and hypothyroidism
were registered as part of the monitoring of a national iodine
supplementation program, before and for 2 yr after iodized
salt became available. However, the market share was low,
4468
J Clin Endocrinol Metab, October 2002, 87(10):4462– 4469
and the average increase in iodine intake was calculated to
be around 5 ␮g/d. The monitoring of two more time periods
confirmed the differences between areas observed during the
initial study period. In addition, a small but statistically
significant increase in the incidence of hyperthyroidism occurred in the area with the lowest iodine intake. Considering
the low market share of iodide containing salt, this increase
in incidence was unexpected. Further observation will reveal
the magnitude and duration of this variation.
The study was observational, based on populations with
long-term stable iodine intake. It may take years from the
time that an increase in iodine intake takes place until a new
steady state in occurrence of hyperthyroidism is reached (31,
32). Initially, there may be a transient increase in the incidence of hyperthyroidism (33, 34), which should then fall to
a lower level. Little is known about the dynamics of the
epidemiology of hypothyroidism after a sudden change in
iodine intake.
The number of patients diagnosed to have hyper- or hypothyroidism in an area may depend on the diagnostic activity. Individuals of the two population cohorts were not
actively investigated for thyroid disease, as they were in the
classical Whickham survey (35). In general, the health care
system and guidelines for investigation of patients were similar in the two areas. The register accumulated information
on the number of thyroid function tests performed in the two
cohorts. There were small differences with slightly more TSH
analyses performed in Copenhagen and more thyroid hormone tests in Aalborg. This difference could not explain the
observed differences in incidence rates of thyroid diseases.
Many tests are used for control of disease, and a major reason
for the small discrepancy in number of tests between areas
may be differences in the abnormalities controlled. New
cases identified in the present study had to seek medical
assistance, and thyroid tests had to be taken. This might
theoretically bias results. However, when we actively examined smaller cohorts from the population of the same areas,
the pattern observed was the same (28).
In conclusion, small differences in iodine intake of populations with mild and moderate ID are associated with
considerable differences in incidence rates of thyroid hyperand hypofunction. The results suggest that small increases in
iodine intake of a population with mild and moderate ID will
prevent a considerable proportion of nodular hyperthyroidism and goiter. On the other hand, the increase in incidence
of thyroid hypofunction observed with an increase in iodine
intake may start already below optimal iodine intake levels.
The optimal level of iodine intake to prevent thyroid disease
may be a relatively narrow range around the recommended
daily intake at 150 ␮g.
Acknowledgments
We are indebted to the GPs in Northern Jutland and Copenhagen and
the involved laboratories for their extraordinary spirit of collaboration.
Received May 14, 2002. Accepted July 2, 2002.
Address all correspondence and requests for reprints to: Inge Bülow
Pedersen, Department of Endocrinology and Medicine, Aalborg Hospital, Reberbansgade, DK-9000 Aalborg, Denmark. E-mail: Ibulow@aas.
nja.dk.
Pedersen et al. • Thyroid Disease and Iodine Intake
This study was a part of the Danish Investigation of Iodine Intake and
Thyroid Diseases (DanThyr), which is supported by grants from the
Danish Medical Foundation, the 1991 Pharmacy Foundation, North
Jutland County Research Foundation, Tømmerhandler Wilhelm Bangs
Foundation, and Copenhagen Hospital Corporation Research
Foundation.
References
1. WHO-FAO-IAEA 1996 Trace elements in human nutrition and health: iodine.
Geneva: World Health Organization; Technical Report 49
2. Hetzel BS, Dunn JT 1989 The iodine deficiency disorders: their nature and
prevention. Annu Rev Nutr 9:21–38
3. Delange F 2000 Iodine deficiency. In: Braverman LE, Utiger RD, eds. The
thyroid, 8th Ed. Philadelphia: Lippincott; 295–316
4. International Council for the Control of Iodine Deficiency Disorders, Information on iodine supplementation: http://www.iccidd.org (view country
information)
5. Pisarev MA, Gärtner R 2000 Autoregulatory actions of iodine. In: Braverman
LE, Utiger RD, eds. The thyroid, 8th Ed. Philadelphia: Lippincott; 85–90
6. Tonacchera M, Agretti P, Chiovato L, Rosellini V, Ceccarini G, Perri A,
Viacava P, Naccarato AG, Miccoli P, Pinchera A, Vitti P 2000 Activating
thyrotropin receptor mutations are present in nonadenomatous hyperfunctioning nodules of toxic or autonomous multinodular goiter. J Clin Endocrinol
Metab 85:2270 –2274
7. Krohn K, Wohlgemuth S, Gerber H, Paschke R 2000 Hot microscopic areas
of iodine-deficient euthyroid goitres contain constitutively activating TSH
receptor mutations. J Pathol 192:37– 42
8. Laurberg P, Pedersen KM, Vestergaard H, Sigurdsson G 1991 High incidence
of multinodular toxic goitre in the elderly population in a low iodine intake
area vs. high incidence of Graves’ disease in the young in a high iodine intake
area: comparative surveys of thyrotoxicosis epidemiology in East-Jutland Denmark and Iceland. J Intern Med 229:415– 420
9. Laurberg P, Pedersen KM, Hreidarsson A, Sigfusson N, Iversen E, Knudsen
P 1998 Iodine intake and the pattern of thyroid disorders: a comparative
epidemiological study of thyroid abnormalities in the elderly in Iceland and
in Jutland, Denmark. J Clin Endocrinol Metab 83:765–769
10. Roti E, Vagenakis AG 2000 Effect of excess iodide: clinical aspects. In: Braverman LE, Utiger RD, eds. The thyroid, 8th Ed. Philadelphia: Lippincott; 316 –329
11. Li M, Liu D, Qu C, Zhang P, Qian Q, Zhang C, Jia Q, Wang H, Eastman CJ,
Boyages SC, Collins JK, Jupp JJ, Maberly GF 1987 Endemic goitre in Central
China caused by excessive iodine intake. Lancet 1:257–259
12. Konno N, Makita H, Yuri K, Iizuka N, Kawasaki K 1994 Association between
dietary iodine intake and prevalence of subclinical hypothyroidism in the
coastal regions of Japan. J Clin Endocrinol Metab 78:393–397
13. Suzuki H, Higuchi T, Sawa K, Ohtaki S, Horiuchi Y 1965 “Endemic coast
goitre” in Hokkaido, Japan. Acta Endocrinol (Copenh) 50:161–176
14. Safran M, Paul TL, Roti E, Braverman LE 1987 Environmental factors affecting
autoimmune thyroid disease. Endocrinol Metab Clin North Am 16:327–342
15. Sundick R, Bagchi N, Brown TR 1992 The role of iodine in thyroid autoimmunity: from chickens to humans. A review. Autoimmunity 13:61– 68
16. Tunbridge WMG, Brewis M, French JM 1981 Natural history of autoimmune
thyroiditis. Br Med J 282:258 –262
17. Brown TR, Bagchi N 1992 The role of iodine in the development of autoimmune thyroiditis. Int Rev Immunol 9:167–182
18. Laurberg P, Bülow Pedersen I, Knudsen N, Ovesen L, Andersen S 2001
Environmental iodine intake affects the type of nonmalignant thyroid disease.
Thyroid 11:457– 469
19. Pedersen KM, Nøhr SB, Laurberg P 1997 Iodine intake in Denmark. Ugeskr
Læger 159:2201–2206
20. Pedersen KM, Laurberg P, Nøhr S, Jørgensen A, Andersen S 1999 Iodine in
drinking water varies by more than 100-fold in Denmark. Importance for
iodine content of infant formulas. Eur J Endocrinol 140:400 – 403
21. 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 Nutr
87:61– 69
22. Knudsen N, Bülow I, Jørgensen T, Laurberg P, Ovesen L, Perrild H 2000
Goitre prevalence and thyroid abnormalities at ultrasonography: a comparative epidemiological study in two regions with slightly different iodine status.
Clin Endocrinol (Oxf) 53:479 – 485
23. Nøhr SB, Laurberg P, Børlum KG, Pedersen KM, Johannesen PL, Damm P,
Fuglsang E, Johansen A 1994 Iodine status in neonates in Denmark: regional
variations and dependency on maternal iodine supplementation. Acta Pædiatr
83:578 –582
24. Pedersen KM, Iversen E, Laurberg P 1995 Urinary iodine excretion and
individual iodine supplementation among elderly subjects: a cross-sectional
investigation in the commune of Randers, Denmark. Eur J Endocrinol 132:
171–174
25. Bülow Pedersen I, Laurberg P, Arnfred T, Knudsen N, Jørgensen T, Perrild
H, Ovesen L 2002 Surveillance of disease frequency in a population by linkage
Pedersen et al. • Thyroid Disease and Iodine Intake
26.
27.
28.
29.
30.
to diagnostic laboratory databases. A system for monitoring the incidences of
hyper- and hypothyroidism as part of the Danish iodine supplementation
program. Comput Methods Programs Biomed 67:209 –216
Clayton D, Hills M 1993 Statistical models in epidemiology. New York: Oxford University Press; 138 –139
Laurberg P, Nøhr SB, Pedersen KM, Hreidarsson A, Andersen S, Bûlow
Pedersen I, Perrild H, Jørgensen T, Ovesen L 2000 Thyroid disorders in mild
iodine deficiency. Thyroid 10:951–963
Knudsen N, Bülow 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. Eur J Endocrinol 143:
485– 491
Shan Z, Teng W, Gao T 2000 The high risk of hypothyroidism in the rural
community of China with high iodine intake after universal salt iodination for
3 years. Endocr J 47(Suppl):P-024
Li Y, Teng W, Guan H, Shan Z, Wang Z 2000 The higher prevalence of
J Clin Endocrinol Metab, October 2002, 87(10):4462– 4469 4469
31.
32.
33.
34.
35.
autoimmune thyroid disease in the rural community with high iodine intake
after 3-year universal salt iodination in China. Endocr J 47(Suppl):P-025
Stewart JC, Vidor GI 1976 Thyrotoxicosis induced by iodine contamination of
food—a common unrecognised condition? Br Med J 1:372–375
Phillips DI, Barker DJ, Rees Smith B, Didcote S, Morgan D 1985 The geographical distribution of thyrotoxicosis in England according to the presence
or absence of TSH-receptor antibodies. Clin Endocrinol (Oxf) 23:283–287
Fradkin J, Wolff J 1983 Iodide-induced thyrotoxicosis. Medicine 62:1–21
Stanbury JB, Ermans AE, Bourdoux P, Todd C, Oken E, Tonglet R, Vidor G,
Braverman LE, Medeiros-Neto G 1998 Iodine-induced hyperthyroidism: occurrence and epidemiology. Thyroid 8:83–100
Vanderpump MPJ, Tunbridge WMG, French JM, Appleton D, Bates D, Clark
F, Grimley Evans J, Hasan DM, Rodgers 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 (Oxf) 43:55– 68