Clinical Endocrinology (2008) 69, 687–696
REVIEW ARTICLE
The incidence of autoimmune thyroid disease: a systematic review
Blackwell Publishing Ltd
of the literature
Anita McGrogan*, Helen E. Seaman†, John W. Wright‡ and Corinne S. de Vries*
*Department of Pharmacy and Pharmacology, University of Bath, †Department of Pharmacoepidemiology, University of Surrey, Guildford
and ‡Department of Diabetes and Endocrinology, Royal Surrey County Hospital, Guildford, UK
Summary
Objective To undertake a systematic review of literature published
between 1980 and 2008 on the incidence of autoimmune thyroid
disease.
Design All relevant papers found through searches of Medline,
EMBASE and ScienceDirect were critically appraised and an assessment
was made of the reliability of the reported incidence data.
Results The reported incidence of autoimmune hypothyroidism
varied between 2·2/100 000/year (males) and 498·4/100 000/year
(females) and for autoimmune hyperthyroidism, incidence ranged
from 0·70/100 000/year (Black males) to 99/100 000/year (Caucasian
females). Higher incidence rates were found in women compared to
men for all types of autoimmune thyroid disease. The majority of
studies included in the review investigated Caucasian populations
mainly from Scandinavia, Spain, the UK and the USA. It is possible
that nonautoimmune cases were included in the incidence rates
reported here, which would give an overestimation in the incidence
rates of autoimmune disease presented.
Conclusion To our knowledge this is the most comprehensive
systematic review of autoimmune thyroid disease conducted in
the past two decades. Studies of incidence of autoimmune thyroid
disease have only been conducted in a small number of mainly
western countries. Our best estimates of the incidence of hypo-
thyroidism is 350/100 000/year in women and 80/100 000/year in
men; the incidence of hyperthyroidism is 80/100 000/year in women
and 8/100 000/year in men.
(Received 10 March 2008; returned for revision 24 March 2008;
finally revised 22 May 2008; accepted 26 May 2008)
Introduction
Endocrine disease of the thyroid may result in either under- or
overactivity of the gland and may be due to congenital factors,
inadequate levels of dietary iodine intake, pregnancy, radiotherapy,
Correspondence: Corinne de Vries, Department of Pharmacy and
Pharmacology, University of Bath, Bath BA2 7AY, UK.
Tel.: +44 (0) 1225 384228; Fax: +44 (0) 1225 386114;
E-mail: c.de-vries@bath.ac.uk
© 2008 The Authors
Journal compilation © 2008 Blackwell Publishing Ltd
doi: 10.1111/j.1365-2265.2008.03338.
viral infection, surgery, underlying disease such as infiltrative disorders,
or autoimmunity.1–5 The incidence of thyroid disease caused by
iatrogenic factors and iodine deficiency is clearly determined by
environmental and clinical factors.6 However, the pathogenesis of
autoimmune thyroid disease is poorly understood although genetic
factors are clearly important in both Graves’ disease and autoimmune
hypothyroidism.7 Insight into the incidence of disease is important
for the identification of trends in relation to patient characteristics
such as sex and geographical location and to determine any changes
in incidence following the introduction of new environmental
factors.
Previous reviews on the epidemiology of thyroid disease have
evaluated the incidence and prevalence of the different types of
thyroid disease, with emphasis being given to the well-known
Whickham study from the UK.1,6,8,9 Other subject areas covered
include the prevalence of subclinical hyper- and hypothyroidism,2 the
epidemiology of hypothyroidism
10,11 11
and thyrotoxicosis. Typically,
the prevalence of hypothyroidism has been found to be between 0 and
7·8/1000 men and between 0 and 20·5/1000 women, and for hyper-
thyroidism, between 2·0 and 19·4/1000 women.11 To our knowledge,
no systematic investigation of the incidence of all types of autoimmune
thyroid disease has been conducted within the past two decades. This
literature review critically appraises relevant published work on the
incidence of primary autoimmune thyroid disease classified clinically
as hypothyroidism, hyperthyroidism, thyrotoxicosis, Graves’ disease
or Hashimoto’s thyroiditis throughout the world.
Method
Searches of the Medline (1980–2008), EMBASE (1980–2008) and
ScienceDirect (1980–2008) databases were conducted using the
keywords ‘hypothyroidism’, ‘hyperthyroidism’, ‘thyrotoxicosis’,
‘Graves disease’, ‘thyroiditis’ and ‘incidence’ or ‘epidemiology’.
The inclusion criteria were that the studies reported original work,
that a reasonable effort had been made by the authors to include all
incident cases, and that the estimates of population size and person-
time contributed were accurate. When assessing the likelihood of
missing incident cases, papers were evaluated as follows. (1) For
case-finding studies, did the authors ensure that all of the subjects
contributing to incidence denominator data would have been
eligible to have the disease diagnosed and did the authors check all
relevant medical records? (2) For all studies, were cases checked to
687
688 A. McGrogan et al.
ensure that they were incident and not prevalent? (3) For all studies,
did the authors ensure that the cause of thyroid disease was
autoimmune and not secondary to another disease or environmental
factors? Where possible, we only included incidence rates for cases
of thyroid disease caused by autoimmunity, determination of which
relied on information given in the paper.
The titles and abstracts of all of the studies produced by the
searches were reviewed by two investigators and those papers
accepted for inclusion in the study were appraised; inclusion of
papers was discussed by three of the authors. Studies published in
English, French, German, Spanish or Dutch were included. Review
papers identified were searched for secondary references reporting
on original research; secondary references found from any of the
other papers reviewed were also included.
A standard data abstraction form was used to record all details of
the papers reviewed; a copy of this is available on request from the
authors. Each study was classified as being at low, medium or high
risk for under- or overestimation of reported incidence rates by
considering the reliability of numerator and denominator data. For
instance, inclusion of prevalent cases or those thought not to be
caused by autoimmunity will have led to overestimated rates, as will
underestimated denominator data. Conversely, missing cases or an
overestimated denominator (e.g. a catchment area from which not
all inhabitants had access to hospital services) would be considered
to result in underestimated incidence rates. Explanations provided
by the papers’ authors as to why incidence rates were as expected or
whether they were considered to be an over- or underestimate of the
true incidence rate were taken into account in this process. If the
extent of likely error was considered to be very great, the study was
excluded. To minimize subjectivity, this assessment was agreed
between the authors and random checks were performed to ensure
consistency. Rates are presented as the number of cases/100 000/
year and, where sufficient data were given in the paper, rates were
checked for accuracy.
Results
The initial database search identified 2227 papers. Of these, 2127
were rejected after reviewing the abstracts. Approximately half of the
studies rejected at abstract review either did not focus on thyroid
disease as the main topic of the paper, reported on thyroid disease
secondary to another disease, or reported on the incidence of certain
diseases among those with thyroid disease. Other key reasons for
excluding papers were that the thyroid disease was caused by drugs,
was congenital or occurred postpartum, treatment of thyroid disease
was the main topic of the paper or the paper covered individual case
studies rather than evaluation of a population. References to those
papers rejected following the abstract review are available on request
from the authors. One hundred full papers were appraised. From
these, 22 further secondary references were identified. Out of the 122
full papers reviewed, 99 were rejected, leaving 23 papers that were
included and critically appraised in this review. The most common
reasons for rejecting papers at this stage were: (1) they were review
papers; (2) they reported incident cases but incidence rates could not
be determined from the papers; (3) they reported on changes in
incidence rates following changes in iodine content of salt or other
Journal compilation © 2008 Blackwell Publishing Ltd, Clinical Endocrinology, 69, 687–696
nonautoimmune thyroid disease; and (4) they reported on prevalence
rather than incidence. Most of the studies included in this review
were conducted among Caucasian populations in Europe; few data
on other ethnic groups were identified.
Hypothyroidism
We identified five studies that reported on the incidence of
hypothyroidism9,12–15 (Table 1) and in all of these, rates were higher
among women than men. All of these studies reported on the
incidence of disease in predominantly Caucasian populations in
Spain,14 Sweden12 and the UK.9,13,15 Three of the studies9,12,14 examined
hospital records using levels of TSH and thyroid hormone as indica-
tors of hypothyroidism; Flynn et al.13 and Leese et al.15 identified
cases of hypothyroidism from a prescription database (using
prescriptions for anti-thyroid and thyroid replacement medication),
a patient index, a thyroid register, morbidity records and a database
with biochemistry data (using ICD-9 and ICD-10 codes). In the
study by Vanderpump et al.,9 patients were asked to complete a
verbal questionnaire, thyroid size was graded and for those who had
died, death certificates were checked as well as general practitioner
(GP), hospital and postmortem records.
The rates presented by Vanderpump et al.,9 Flynn et al.13 and Leese
15 9,13 12
et al. for all ages compared well. Sundbeck et al. reported
incidence only in females aged 70–81 years whereas Galofre et al.14
reported the lowest rates, which are thought to underestimate the
true incidence rates.
Hyperthyroidism
The incidence rates for these conditions have been reported on
together and are shown in Table 2. Rates were given for predominantly
Caucasian populations; only one study16 gave comparisons of
incidence rates between different ethnic groups. They found that
incidence rates in Asian and ‘coloured’ (terminology adopted from
original paper) populations in Johannesburg, South Africa were
slightly higher than the Caucasian populations but incidence was
lowest in the African population. Five studies reported on incidence
in children and found, overall, that rates were between 0·1 and 5·0/
100 000/year.
The majority of studies based the diagnosis on both clinical
criteria and biochemical tests of thyroid function.3,14,16–29 Cases were
identified for these studies either through sending questionnaires to
relevant departments3,20,21,29,30 or by checking hospital medical
14,16–19,22–28 31
records or test result databases. One study used the
Oakland Kaiser-Permanente Medical plan and free T4 index tests
ordered by physicians to locate cases. One study did not give details
about diagnostic criteria used;32 all of the studies looking at hypo-
thyroidism also reported on hyperthyroidism, the details of which
have been given above.9,12–15
The incidence of overactive thyroid disease in Caucasian males
ranged from 2·1/100 000/year in Spain14 to 22·0/100 000/year15 in
Scotland, and in Caucasian females from 23·4/100 000/year in New
Zealand3 to 99/100 000/year in Scotland.15 Sundbeck et al.12 found a
slightly higher incidence rate in 70–81-year-old females but did
not look at any other patient age groups. Where incidence rates
© 2008 The Authors
 Incidence of autoimmune thyroid disease 689

Table 1. The incidence of hypothyroidism by age and sex (crude rates/100 000/year unless stated otherwise)

Rate/100 000/year (95% CI, if available)

Age
Study Location Period (years) Diagnostic test/criteria Males Females Both

Population-based studies
Flynn et al.13 Scotland 1993–1996 All Long-term TRT. No 87·6 (80·0–95·3) 498·4 (17·9–480·6) 297·5
Study: Population registered separate identification (287·6–307·3)
with GP. Cases identified from of AI TD
six databases, including patient
records, laboratory data,
prescriptions, thyroid register
Risk of underestimation**
Risk of overestimation*
Leese et al.15 Scotland 1997 All As above 82 (70–96) 434 (406–463)
Study: As above 1998 76 (64–89) 396 (369–424)
Risk of underestimation** 1999 105 (91–120) 466 (437–497)
Risk of overestimation* 2000 111 (97–127) 424 (396–453)
2001 109 (95–125) 475 (446–507)
9
Vanderpump et al. UK 1972–1993 All TSH; Tg Ab, TMS Ab and 80† 250†
Study: Prospective, sample of anti-thyroid cytoplasmic 60 (30–120)‡§ 350 (280–450)‡
population followed up antibodies; FT4; thyroid size.
Risk of underestimation* Identification of AI TD
Risk of overestimation* from Ab tests

Population sample studies

Galofre et al.14 Viga, Spain 1990–1992 All Hypothyroid status, positive 2·2 (0–6·5) 45·4 (28·0–62·9) 26·2
Study: Studied a random sample for Tg Ab or TMS Ab. (16·3–36·1)
of outpatients (~10% of city Identification
population) who met the of AI TD from Ab tests
diagnosis criteria as found
on laboratory registers
Risk of underestimation***
Risk of overestimation*
12
Sundbeck et al. Sweden 1971–1988 70–81 TSH. No separate 200
Study: Prospective, sample of identification of AI TD
population followed up
Risk of underestimation**
Risk of overestimation*

Risk of under or overestimation: *low, **medium, ***high; †cases/100 000 deceased/year; ‡cases/100 000 survivors/year; §includes one subject whose
hypothyroidism was caused by lithium therapy; FT4, free T4; Tg Ab, thyroglobulin antibody; TMS Ab, thyroid microsomal antibody; TRT, thyroid replacement
therapy; AI, autoimmune; TD, thyroid disease; Ab, antibody.

were presented for males and females together, these ranged

from 5·6/100 000/year in Serbia28 to 52/100 000/year in Oakland,
31
USA.
Comparing the incidence rates between different geographical
locations within the countries reported on is difficult because only
four studies gave rates by different locations3,18,19,28 and none of these
reported fundamental differences in incidence rates between the
locations studied. Similarly, once study method had been taken into
account, no large differences in incidence rates were observed
between different countries. Where possible, only incidence rates of
thyroid disease caused by autoimmunity were included; however,
some studies did not give sufficient detail for this determination to
be made and so nonautoimmune cases of thyroid disease may have
been included in some of the rates.
Discussion
Incidence rates of autoimmune hypothyroidism varied between
2·2/100 000/year and 498·4/100 000/year whereas incidence rates of
autoimmune hyperthyroidism varied between 0·7/100 000/year and
99/100 000/year. Thyroid disease had a higher incidence in women.
In their review of the epidemiology of thyroid diseases, Tunbridge
and Caldwell6 point out that complications arise due to problems of
definition, selection criteria and different techniques used for the
measurement of thyroid function. In addition, symptoms of thyroid
disease may be nonspecific or attributed to other diseases, which
makes diagnosis more difficult.33
When thyroid disease is caused by environmental factors, such as
levels of iodine, incidence rates have been found to vary between

© 2008 The Authors

Journal compilation © 2008 Blackwell Publishing Ltd, Clinical Endocrinology, 69, 687–696
 Table 2. The incidence of diseases of the overactive thyroid (crude rates/100 000/year unless stated otherwise)

690 A. McGrogan et al.

Rate /100 000/year (95% CI, if available)
Age
Study Location Period Diagnostic test/criteria (years) Males Females Both

Population-based studies
Flynn et al.13 Tayside, Scotland 1993–1997 Received treatment for All 14·3 (11·4–17·2) 76·9 (70·3–83·5) 46·3 (42·7–49·9)
Study: Population registered with GP. hyperthyroidism. No separate
Cases identified from six databases: identification of AI TD
including patient records, laboratory
data, prescriptions, thyroid register
Risk of underestimation**
Risk of overestimation*
Leese et al.15 Tayside, Scotland 1997 As above All 13 (9–20) 77 (65–90)
Study: As above 1998 16 (11–23) 78 (66–91)
Risk of underestimation** 1999 18 (12–25) 95 (82–109)
Risk of overestimation* 2000 22 (16–30) 99 (85–131)
2001 15 (10–22) 91 (78–105)
9
Vanderpump et al. North-east, UK 1972–1993 TSH; Tg Ab, TMS Ab and anti-thyroid All 120§
Journal compilation © 2008 Blackwell Publishing Ltd, Clinical Endocrinology, 69, 687–696

Study: Prospective, sample cytoplasmic antibodies; FT4; thyroid size. 80 (60–140)¶

of population followed up Identification of AI TD from Ab tests
Risk of underestimation*
Risk of overestimation*

Case-finding studies
Wong and Cheng23 Hong Kong, China 1989–1998 Clinical features; size of thyroid; FT4; T3; 0–14 (Asian) 0·9 (0·0–5·0) 9·5 (4·8–17·1) 5·0 (2·6–8·8)
Study: Registry and review of patient TSH; TRAb from 1992. Identification of
database; all cases referred were AI TD from Ab tests from 1992
assessed in the Endocrine clinic
Risk of underestimation*
Risk of overestimation*
Wong et al.24 Hong Kong, China 1990–1994 0–14 (Asian) 0·8 (0·0–4·4) 7·2 (3·1–14·1) 3·8 (1·8–7·3)
As above
Forssberg et al.22 Oredbro, Sörmland, 1990–1999 TFT; TSH; FT4; clinical symptoms; 0–16 5·0 2·9 1·7
Study: Diagnostic registers reviewed Västmanland and immunological markers. Identification
to find cases; no information given Darlarna, Sweden of AI TD from Ab tests
about who referred patients for tests
Risk of underestimation*
Risk of overestimation*
© 2008 The Authors

27
Berglund et al. Malmö, Sweden 1988–1990 Clinical signs; serum T3, T4 and TSH. All 8·8 34·4 22·2
Study: Hospital records of diagnoses No separate identification of AI TD
and treatment; no information about
who referred patients for tests
Risk of underestimation*
Risk of overestimation*
Journal compilation © 2008 Blackwell Publishing Ltd, Clinical Endocrinology, 69, 687–696
© 2008 The Authors

Table 2. Continued

Rate /100 000/year (95% CI, if available)

Age
Study Location Period Diagnostic test/criteria (years) Males Females Both

Berglund et al.17 Malmö, Sweden 1970–1974 Clinical symptoms; serum PBI; T3; All 7·4 27·2 17·7
Study: Hospital records of diagnoses presence/absence of TRAb; TRI uptake.
and treatment; no information about Identification of AI TD from Ab tests
who referred patients for tests
Risk of underestimation*
Risk of overestimation*
Cox et al.25 Stoke-on-Trent, 1974–1983 Treated for HT; if levels of FT4 index or All 5·5‡ 25·8‡
Study: Records of TFTs; case notes also Newcastle-under-Lyme FT4 exceeded reference range or diagnosed
checked; no information given on who and Leek, England with GD and had ophthalmopathy, diffuse
referred patients thyroid scan or raised TMS Ab. Unclear
Risk of underestimation* whether AI TD was identified in all cases
Risk of overestimation*
Barker and Phillips18 12 towns, 1982 TFT. No separate identification of AI TD All 9·2‡ 35·5‡ 22·7‡
Study: Cases identified from laboratory England and Wales
test registers. No information about
who referred patients for tests
Risk of underestimation*
Risk of overestimation*
Kalk and Kalk26 Johannesburg, 1974–1984 Usual clinical criteria, elevated FT4, or by All
Study: Records of those referred for South Africa (excluding checking FT4 and T3 levels. Diffuse goitre African 0·70† 8·75†
thyroid scans; no information about 1976) confirmed by concentration of the isotope
who referred patients throughout gland. No separate identification
Risk of underestimation** of AI TD

Incidence of autoimmune thyroid disease 691

Risk of overestimation*
Kalk16 Johannesburg, 1974–1975, Usual clinical criteria; elevated FT4; 24-h All:
As above South Africa 1977–1978 radioiodine or technetium uptake tests. African 1·9 23
No separate identification of AI TD Coloured 6·9 51
Asian White 9·6 48
8·4 42
Lavard et al.20 Denmark 1982–1988 Appearance, T4, T3, TSH, T3, TFT, TRIs. 0–4 0·1 0·1
Study: Questionnaires to hospital No separate identification of AI TD 5–9 0·1 1·0
departments asking for information 10–14 0·5 3·0
on cases diagnosed; checked hospital
admission records
Risk of underestimation**
Risk of overestimation*
 692 A. McGrogan et al.
Table 2. Continued

Rate /100 000/year (95% CI, if available)

Age
Study Location Period Diagnostic test/criteria (years) Males Females Both

Peter30 Budapest, Hungary 1966–1970, FT4, T4, T3 RIA. No separate 0–16 1·2
Study: Questionnaires to hospital 1980–1984 identification of AI TD
departments about newly diagnosed
patients with thyroid disease
Risk of underestimation**
Risk of overestimation*
3
Brownlie and Wells North and 1983–1985 TSH; serum T4; FT4; serum T3; TMS Ab; All 23·5 (20·8–26·3)
Study: Cases identified in thyroid clinic mid-Canterbury, Tg Ab. Identification of AI TD from Ab tests
records; no information on who New Zealand
referred patients to clinic Questionnaires
sent to GPs of patients diagnosed
Journal compilation © 2008 Blackwell Publishing Ltd, Clinical Endocrinology, 69, 687–696

Risk of underestimation**
Risk of overestimation**
21
Haraldsson et al. Iceland 1980–1982 RIA; T3; T4; free T4; resin T3; TRH All 8·9 38·4 23·6
Study: Questionnaires to GPs, response; T3 suppress; thyroid Ab tests;
enquiries made to hospitals and occasional thyroid scan. Identification
Endocrine Society members requesting of AI TD from Ab tests
information on incident cases
Risk of underestimation**
Risk of overestimation**
Mogensen and Green19 Funen, Denmark 1972–1974 TFT: serum T4 and T3; resin T3; All 8·7 (6·6–10·8) 46·5 (43·3–49·7)
Study: Reviewed all TFT; enquiry to serum PBI. No separate identification
check diagnosis with referring physician of AI TD
Risk of underestimation**
Risk of overestimation**
Paunkovic et al.28 Timok Region, 1971–1990 Clinical, laboratory findings using RIA. All
Study: Continuous registration of cases Serbia Most had 131I and 99mTc thyroid uptake 1971–1980 5·6‡
in the Endocrinology–Thyroid unit and thyroid scintiscanning. From 1986 1981–1990 11·7‡
Risk of underestimation*** Zajecar County, TRAb tested for identification of AI TD 1971–1080 8·0‡
Risk of overestimation** Serbia 1981–1990 15·8‡
Houston et al.32 Midwestern 1954–1983 Diagnosed with hyperthyroidism and All 25·9
Study: No details given community, USA living in Rochester during study.
© 2008 The Authors

Risk of underestimation** No separate identification of AI TD

Risk of overestimation**
Journal compilation © 2008 Blackwell Publishing Ltd, Clinical Endocrinology, 69, 687–696
© 2008 The Authors

Table 2. Continued

Rate /100 000/year (95% CI, if available)

Age
Study Location Period Diagnostic test/criteria (years) Males Females Both

Holm et al.29 USA 1989–2001 Defined as definite or probable.‡‡ All 460††

Study: Questionnaire to Nurses’ Health No separate identification of AI TD
Study II asking about diagnosis; cases
followed up with enquiry to physician
Risk of underestimation**
Risk of overestimation**
dos Remedios et al.31 Oakland, USA 1976 FT4 index. No separate All 52
Study: Oakland Kaiser-Permanente 1977 identification of AI TD 46
medical plan patients; thyroid tests
requested in symptomatic patients
by medical centre physicians
Risk of underestimation*
Risk of overestimation***

Population sample
Galofre et al.14 Viga, Spain 1990–1992 Altered level of at least one of: TSH, FT4, All 2·1 (0·0–6·4) 41·9 (25·0–58·8) 24·2 (14·7–33·7)
Study: Studied a random sample of Tg Ab and TMS Ab; positive TRAb. Overt 22·3 (13·2–31·4)
outpatients (~10% of city population) Identification of AI TD from Ab tests Subclinical 1·9 (0–4·6)
who met the diagnosis criteria as found
on laboratory registers
Risk of underestimation***
Risk of overestimation*

Incidence of autoimmune thyroid disease 693

Sundbeck et al.12 Gothenburg, 1971–1988 TSH. No separate identification 70–81 100
Study: Prospective, sample of Sweden of AI TD
population followed up
Risk of underestimation**
Risk of overestimation***

Risk of under- or overestimation: *low, **medium, ***high; †figure read from graph; ‡age- and sex-adjusted incidence rate; §cases/100 000 deceased/year; ¶cases/100 000 survivors/year; TFT, thyroid function
tests; TRH, thyrotrophin-releasing hormone; RIA, radioimmunoassay; FT4, free T4; TRAb, thyrotrophin receptor antibody; Tg Ab, thyroglobulin antibody; TMS Ab, thyroid microsomal antibody; PBI, protein-
bound iodine; TRI, thyroid radioiodine; TRIs, thyroid radioisotope scintigram; GD, Graves’ disease; HT, hyperthyroidism; AI, autoimmune; TD, thyroid disease; Ab, antibody; ††indicates the rate where the units
given are possibly incorrect by a factor of 10; therefore, the rate is included in the table for completeness but has been excluded from discussion. ‡‡Definite: high T4, low TSH, high TRI uptake or diffuse pattern
on scan, treated for HT, absence of other thyroid disorders. ‘Probable’: at least one symptom of HT or one of goitre or eye changes, treated for HT and absence of other thyroid disorder; or who had all of the
clinical signs and consistent laboratory test results but were not treated.
694 A. McGrogan et al.
locations and over time.34–37 For instance, in their study of the
prevalence of thyroid disease in the elderly, Laurberg et al.38 found
a high prevalence of hypothyroidism in Iceland, where the intake of
iodine was high, but in Jutland, where iodine intake was low, a
high prevalence of hyperthyroidism was found. In this review of
autoimmune thyroid disease, the papers we identified came from a
limited range of geographical areas. Consequently we could not
comment on the absence or presence of differences in incidence rates
between different geographical locations.
There were over 100-fold differences in the incidence rates of
various studies. Higher incidence rates tended to be due to the type
of study conducted (e.g. where screening was used and therefore
subclinical cases were included in the rates). The two prospective
studies9,12 produced the highest incidence rates of thyroid disease.
14
One of these only included women aged 70–81 years; the other
used a multitude of data sources to evaluate and screen incident cases
in the entire population and is the most comprehensive study we
identified. The incidence rates of hypothyroidism in this study,
which was carried out in the UK, were between 250 and 350/100 000/
year (depending on the subgroup of the population). This was
among the higher rates identified, partly as a result of the inclusion
of subclinical cases. By contrast, a study carried out in Spain using
a selected outpatients list reported an incidence rate in women of
45·4/100 000/year for hypothyroid disease; they will have missed any
cases of thyroid disease who did not present at the participating
outpatient clinic. In general, the incidence rates identified in the
prospective studies will be more accurate. The difficulty with using
these rates for post hoc evaluation of changes in incidence rates is
that the prospective studies will have included subclinical cases.
Looking at the results for hyperthyroidism from retrospective
studies, it is useful to note that the studies conducted in a similar
way, through case finding from questionnaires, medical records or
test results,3,14,18,19,21,27 produced similar incidence rates even though
these studies covered different time periods between 1972 and 1999.
This is an important finding as it indicates that, for autoimmune
thyroid disease, the rates appear to be constant over time. However,
a recent study from Scotland15 found that the incidence of
hyperthyroidism in females and hypothyroidism in males increased
between 1997 and 2001. The authors note that this may be partly
explained by an increase in the number of thyroid tests being carried
out in the region, leading to an increased number of subclinical cases
being identified. If this were a correct assumption then increases in
the incidence rates for both types of thyroid disease and in both
males and females would be expected unless there was differential
testing between males and females. A true increase in incidence of
thyroid disease caused by autoimmunity or some other cause cannot
be ruled out but it is also possible that the increase seen was caused
by an artefact. The studies included in this review mostly covered
Caucasian populations, therefore we are unable to comment on
potential differences in incidence rates between different ethnic
groups.
In reviews covering the epidemiology of thyroid disorders,
the distinction has been made between subclinical and overt hypo-
thyroidism and hyperthyroidism.1,2,6,11 In this review we have found
only two studies that were conducted in a way that would
include both subclinical and overt cases of hypothyroidism or
Journal compilation © 2008 Blackwell Publishing Ltd, Clinical Endocrinology, 69, 687–696
hyperthyroidism: Vanderpump et al.9 and Sundbeck et al.12 examined
all patients in their study cohorts for thyroid disease, which allowed
them to detect subclinical cases of hypothyroidism and hyper-
thyroidism. All other studies used methods of case finding that did
not involve the screening of patients. As Vanderpump et al. pointed
out, cases will be missed unless patients are screened for thyroid
disease;9 except for incidental findings subclinical disease will not, by
definition, be readily detected clinically. In addition, the point where
a patient is diagnosed and treated for thyroid disease, and when their
disease becomes overt, differs widely in clinical practice and this will
have resulted in differences in incidence rates between different
geographical locations. However, given the combination of differences
in rates and study design in different geographical locations, we cannot
exclude the possibility that there is a geographical component to
variations in incidence of thyroid disease.
Only two prospective and truly population-based studies9,13 were
identified. It could be argued that studies that were not population
based or collected data retrospectively are more prone to producing
under- or overestimated rates. For example, retrospective studies rely
more heavily on physician or patient recall and will not necessarily
identify all patients if records were destroyed, sent on to different
hospitals, or otherwise lost for research purposes. These instances
will have led to underestimated incidence rates. Inclusion of non-
autoimmune disease (which may be more difficult to establish
retrospectively) will have led to overestimated rates. Finally, retro-
spective assessment of population size or person-time contributed
is often more difficult and therefore more prone to error than
prospective collection of this information.
In an assessment of the incidence of autoimmune thyroid disease
another important consideration is the likely cause of thyroid
disease. Hypothyroidism may be caused by other factors including
the exogenous causes of medication with lithium, radioiodine or
anti-thyroid drugs and thyroidectomy as well as the endogenous
cause of autoimmune thyroiditis.1,2 Similarly, in addition to auto-
immune Graves’ disease, hyperthyroidism may be caused by overzealous
thyroid hormone replacement therapy, or by other endogenous
thyroid disease including acute viral thyroiditis, toxic multinodular
goitre or an autonomous adenoma. Rare causes include struma
ovarii and pituitary hypersection of TSH.2 In this report, where
possible, only incidence rates for autoimmune causes of thyroid
disease have been included, although most papers did not specify in
detail the causes of the thyroid disease reported. Associated with this
issue is the possibility that euthyroid subjects with thyroid antibodies
in the serum were included in incidence rates in those retrospective
study designs that solely used results of biochemical tests for the
diagnosis of thyroid disease. Therefore, it is possible that the rates
presented overestimate the incidence of autoimmune thyroid
disease.
The majority of studies included were conducted in the USA,
Scandinavia and the UK. The absence of reports on incidence rates
of autoimmune thyroid disease from the rest of the world is striking
and possibly reflects the more substantial problem of iodine-related
thyroid disease in these other regions. Endemic goitre affects 200
million people throughout the world, particularly in the Himalayas,
Andes, parts of Africa and eastern and southern Europe.4 Areas
where iodine prophylaxis has been introduced more recently include
© 2008 The Authors

Poland,35 Austria39 and Switzerland;40 studies of thyroid disorders
in these countries tend to focus on iodine-related rather than
autoimmune thyroid disease.
This review of incidence of thyroid disease has highlighted
difficulties in ensuring that all cases and only those cases of thyroid
disease caused by autoimmunity, whether subclinical or overt, are
included in the incidence rates. The most accurate incidence rates
available come from prospective studies that screen patients.
However, it is thought that retrospective case-finding procedures
produce useful estimations of incidence rates providing the above
limitations are taken into account.
Conclusion
From this comprehensive systematic review of autoimmune thyroid
disease, our best estimate of incidence rates for hypothyroidism
in females is 350/100 000/year and in males, 80/100 000/year;
for hyperthyroidism in females 80/100 000/year and in males, 8/
100 000/year.
Rates were generally higher among females: the incidence of
hypothyroidism was between 2·2 and 111/100 000/year in males and
between 200 and 498·4/100 000/year in females. For disorders of
overactive thyroid disease, incidence rates for overt cases were
between 0·7 and 22·0/100 000/year in males and between 8·8 and
46·5/100 000/year in females; studies including subclinical cases did
not report incidence rates in males but reported higher incidence
rates in females of around 80/100 000/year.
However, in view not only of the limited number of geographical
areas covered but also the differences in methods used to determine
incidence rates, caution is required in applying these figures to
populations elsewhere in the world.
Acknowledgements
This work was supported by a grant from GSK Biologicals.
References
1 Vanderpump, M.P. & Tunbridge, F. (2002) Epidemiology and
prevention of clinical and subclinical hypothyroidism. Thyroid, 12,
839–847.
2 Wiersinga, W.M. (1995) Subclinical hypothyroidism and hyper-
thyroidism. I. Prevalence and clinical relevance. Netherlands Journal
of Medicine, 46, 197–204.
3 Brownlie, B.E. & Wells, J.E. (1990) The epidemiology of thyro-
toxicosis in New Zealand: incidence and geographical distribution
in north Canterbury, 1983–1985. Clinical Endocrinology, 33, 249–
259.
4 Warrell, D.A., Cox, T.M., Firth, J.D. & Benz, E.J. (2003) Oxford Text-
book of Medicine. Oxford University Press, Oxford.
5 Kumar, P. & Clark, M. (1998) Clinical Medicine. W.B. Saunders,
Edinburgh.
6 Tunbridge, W.M. & Caldwell, G. (1991) The epidemiology of
thyroid diseases. In: L.E. Braverman, R.D. Utiger eds. The Thyroid:
A Fundamental and Clinical Text. Lippincott, Philadelphia.
7 Jacobson, E.M. & Tomer, Y. (2007) The genetic basis of thyroid
autoimmunity. Thyroid, 17, 1–13.
© 2008 The Authors
Journal compilation © 2008 Blackwell Publishing Ltd, Clinical Endocrinology, 69, 687–696
Incidence of autoimmune thyroid disease 695
8 Tunbridge, W.M.G., Evered, D.C., Hall, R., Appleton, D., Brewis, M.,
Clark, F., Evans, J.G., Young, E., Bird, T. & Smith, P.A. (1977) The
spectrum of thyroid disease in a community: the Whickham survey.
Clinical Endocrinology, 7, 481–493.
9 Vanderpump, M.P., Tunbridge, W.M., French, J.M., Appleton, D.,
Bates, D., Clark, F., Grimley, E.J., Hasan, D.M., Rodgers, H. &
Tunbridge, F. (1995) The incidence of thyroid disorders in the
community: a twenty-year follow-up of the Whickham Survey.
Clinical Endocrinology, 43, 55–68.
10 Bilous, R.W. & Tunbridge, W.M. (1988) The epidemiology of
hypothyroidism – an update. Bailliere’s Clinical Endocrinology and
Metabolism, 2, 531–540.
11 Vanderpump, M.P. & Tunbridge, W.M. (1996) The epidemiology of
thyroid disease. In: L.E. Braverman, R.D. Utiger eds. The Thyroid.
Lippincott-Raven, Philadelphia. 474–482.
12 Sundbeck, G., Lundberg, P.A., Lindstedt, G., Jagenburg, R. & Eden,
S. (1991) Incidence and prevalence of thyroid disease in elderly
women: results from the longitudinal population study of elderly
people in Gothenburg, Sweden. Age and Aging, 20, 291–298.
13 Flynn, R.W., MacDonald, T.M., Morris, A.D., Jung, R.T. & Leese, G.P.
(2004) The Thyroid Epidemiology, Audit, and Research Study:
thyroid dysfunction in the general population. Journal of Clinical
Endocrinology and Metabolism, 89, 3879–3884.
14 Galofre, J.C., Garcia-Mayor, R.V., Fluiters, E., Fernandez-Calvet, L.,
Rego, A., Paramo, C. & Andrade, M.A. (1994) Incidence of different
forms of thyroid dysfunction and its degrees in an iodine sufficient
area. Thyroidology, 6, 49–54.
15 Leese, G.P., Flynn, R.V., Jung, R.T., MacDonald, T.M., Murphy, M.J.
& Morris, A.D. (2008) Increasing prevalence and incidence of
thyroid disease in Tayside, Scotland: The Thyroid Epidemiology,
Audit and Research Study (TEARS). Clinical Endocrinology, 68, 311–
316.
16 Kalk, W.J. (1981) Thyrotoxicosis in urban black Africans: a rising
incidence. East African Medical Journal, 58, 109–116.
17 Berglund, J., Christensen, S.B. & Hallengren, B. (1990) Total and
age-specific incidence of Graves’ thyrotoxicosis, toxic nodular goitre
and solitary toxic adenoma in Malmo 1970–1974. Journal of Internal
Medicine, 227, 137–141.
18 Barker, D.J. & Phillips, D.I. (1984) Current incidence of thy-
rotoxicosis and past prevalence of goitre in 12 British towns. Lancet,
2, 567–570.
19 Mogensen, E.F. & Green, A. (1980) The epidemiology of thyrotoxi-
cosis in Denmark. Incidence and geographical variation in the Funen
region 1972–1974. Acta Medica Scandinavica, 208, 183–186.
20 Lavard, L., Ranlov, I., Perrild, H., Andersen, O. & Jacobsen, B.B.
(1994) Incidence of juvenile thyrotoxicosis in Denmark, 1982–1988.
A nationwide study. European Journal of Endocrinology, 130, 565–
568.
21 Haraldsson, A., Gudmundsson, S.T., Larusson, G. & Sigurdsson, G.
(1985) Thyrotoxicosis in Iceland 1980–1982. An epidemiological
survey. Acta Medica Scandinavica, 217, 253–258.
22 Forssberg, M., Arvidsson, C.G., Engvall, J., Lindblad, C., Snellman, K.
& Aman, J. (2004) Increasing incidence of childhood thyrotoxicosis
in a population-based area of central Sweden. Acta Paediatrica, 93,
25–29.
23 Wong, G.W. & Cheng, P.S. (2001) Increasing incidence of childhood
Graves’ disease in Hong Kong: a follow-up study. Clinical Endocrino-
logy, 54, 547–550.
24 Wong, G.W., Kwok, M.Y. & Ou, Y. (1995) High incidence of juvenile
Graves’ disease in Hong Kong. Clinical Endocrinology, 43, 697–
700.
696 A. McGrogan et al.
25 Cox, S.P., Phillips, D.I. & Osmond, C. (1989) Does infection initiate
Graves disease? A population based 10 year study. Autoimmunity, 4,
43–49.
26 Kalk, W.J. & Kalk, J. (1989) Incidence and causes of hyperthyroidism
in blacks. South African Medical Journal, 75, 114–117.
27 Berglund, J., Ericsson, U.B. & Hallengren, B. (1996) Increased
incidence of thyrotoxicosis in Malmo during the years 1988–1990 as
compared to the years 1970–1974. Journal of Internal Medicine, 239,
57–62.
28 Paunkovic, N., Paunkovic, J., Pavlovic, O. & Paunovic, Z. (1998) The
significant increase in incidence of Graves’ disease in eastern Serbia
during the civil war in the former Yugoslavia (1992 to 1995). Thyroid,
8, 37–41.
29 Holm, I.A., Manson, J.E., Michels, K.B., Alexander, E.K., Willett,
W.C. & Utiger, R.D. (2005) Smoking and other lifestyle factors and
the risk of Graves’ hyperthyroidism. Archives of Internal Medicine,
165, 1606–1611.
30 Peter, F. (1986) Juvenile thyrotoxicosis. Acta Endocrinologica.
Supplementum, 279, 361–366.
31 dos Remedios, L.V., Weber, P.M., Feldman, R., Schurr, D.A. & Tsoi,
T.G. (1980) Detecting unsuspected thyroid dysfunction by the free
thyroxine index. Archives of Internal Medicine, 140, 1045–1049.
32 Houston, M.S., Hay, I.D. & McConahey, W.M. (1987) Incidence and
etiology of thyrotoxicosis in a midwest community. Clinical Research,
35, 396A.
33 Beastall, G.H., Beckett, G.J., Franklyn, J., Fraser, W.D., Hickey, J.,
John, R., Kendall-Taylor, P., Nevens, B. & Vanderpump, M.P. (2006)
UK Guidelines for the Use of Thyroid Function Tests. Available at:
http://www.british-thyroid-association.org/info-for-patients/Docs/
TFT_guideline_final_version_July_2006.pdf (accessed 13 August 2008).
34 Burgi, H., Kohler, M. & Morselli, B. (1998) Thyrotoxicosis incidence
in Switzerland and benefit of improved iodine supply. Lancet, 352,
1034.
Journal compilation © 2008 Blackwell Publishing Ltd, Clinical Endocrinology, 69, 687–696
35 Lewinski, A., Szybinski, Z., Bandurska-Stankiewicz, E., Grzywa, M.,
Karwowska, A., Kinalska, I., Kowalska, A., Makarewicz, J., Nauman,
J., Slowinska-Klencka, D., Sowinski, J., Syrenicz, A., Zonenberg, A.,
Huszno, B. & Klencki, M. (2003) Iodine-induced hyperthyroidism
– an epidemiological survey several years after institution of iodine
prophylaxis in Poland. Journal of Endocrinological Investigation, 26,
57–62.
36 Laurberg, P., Pedersen, K.M., 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. Journal of Internal Medicine, 229, 415–420.
37 Stanbury, J.B., Ermans, A.E., Bourdoux, P., Todd, C., Oken, E.,
Tonglet, R., Vidor, G., Braverman, L.E. & Medeiros-Neto, G. (1998)
Iodine-induced hyperthyroidism: occurrence and epidemiology.
Thyroid, 8, 83–100.
38 Laurberg, P., Pedersen, K.M., Hreidarsson, A.B., Sigfusson, N.,
Iverson, E. & Knudsen, P.R. (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.
Journal of Clinical Endocrinology and Metabolism, 83, 765–769.
39 Mostbeck, A., Galvan, G., Bauer, P., Eber, O., Atefie, K., Dam, K.,
Feichtinger, H., Fritzsche, H., Haydl, H., Kohn, H., Konig, B.,
Koriska, K., Kroiss, A., Lind, P., Markt, B., Maschek, W., Pesl, H.,
Ramschak-Schwarzer, S., Riccabona, G., Stockhammer, M. &
Zechmann, W. (1998) The incidence of hyperthyroidism in Austria
from 1987 to 1995 before and after an increase in salt iodization in
1990. European Journal of Nuclear Medicine, 25, 367–374.
40 Baltisberger, B.L., Minder, C.E. & Burgi, H. (1995) Decrease of
incidence of toxic nodular goitre in a region of Switzerland after
full correction of mild iodine deficiency. European Journal of
Endocrinology, 132, 546–549.
© 2008 The Authors