A strong familiality of Ankylosing Spondylitis through several generations

Abstract

To elucidate the familiality of ankylosing spondylitis (AS) in Iceland.
The Icelandic genealogy database and population-wide data on all living Icelanders diagnosed as having AS (n=280), who previously had taken part in an epidemiological study on the prevalence of AS in Iceland, were included in the study. Identification of all interpatient relationships in the genealogy database allowed calculation of estimates of the RR for AS in the first-degree relatives (FDRs) to fourth-degree relatives of patients. For each AS proband, 1000 sets of matched Icelandic subjects in the genealogy database were used as controls.
FDRs, second-degree and third-degree relatives had RRs of 75.5, 20.2 and 3.5, respectively (all p values <0.0001), indicating a significantly increased risk for relatives of the patients with AS to develop AS, suggesting a strong heritable factor, while the fourth-degree relatives had a RR of 1.04 (p=0.476) for having AS.
Patients with AS in Iceland are significantly more related to each other than to randomly sampled control subjects. This is in agreement with previous reports on the familiality of AS, but the present study has more power and extends over larger familiar cohorts than previously reported.

Full text

doi: 10.1136/ard.2009.125914
2010 69: 1346-1348 originally published online May 6, 2010Ann Rheum Dis
Arni Jon Geirsson, Kristleifur Kristjansson and Bjorn Gudbjornsson
through several generations
A strong familiality of ankylosing spondylitis
http://ard.bmj.com/content/69/7/1346.full.html
Updated information and services can be found at:
These include:
References http://ard.bmj.com/content/69/7/1346.full.html#ref-list-1
This article cites 19 articles, 2 of which can be accessed free at:
service
Email alerting box at the top right corner of the online article.
Receive free email alerts when new articles cite this article. Sign up in the
Notes
http://ard.bmj.com/cgi/reprintform
To order reprints of this article go to:
http://ard.bmj.com/subscriptions go to: Annals of the Rheumatic DiseasesTo subscribe to
group.bmj.com on June 21, 2010 - Published by ard.bmj.comDownloaded from

Concise report
Ann Rheum Dis 2010;69:1346–1348. doi:10.1136/ard.2009.1259141346
ABSTRACT
Objective To elucidate the familiality of ankylosing
spondylitis (AS) in Iceland.
Methods The Icelandic genealogy database and
population-wide data on all living Icelanders diagnosed
as having AS (n=280), who previously had taken part
in an epidemiological study on the prevalence of AS in
Iceland, were included in the study. Identifi cation of all
interpatient relationships in the genealogy database
allowed calculation of estimates of the RR for AS in the
fi rst-degree relatives (FDRs) to fourth-degree relatives
of patients. For each AS proband, 1000 sets of matched
Icelandic subjects in the genealogy database were used
as controls.
Results FDRs, second-degree and third-degree relatives
had RRs of 75.5, 20.2 and 3.5, respectively (all p values
<0.0001), indicating a signifi cantly increased risk for
relatives of the patients with AS to develop AS, suggesting
a strong heritable factor, while the fourth-degree relatives
had a RR of 1.04 (p=0.476) for having AS.
Conclusions Patients with AS in Iceland are signifi cantly
more related to each other than to randomly sampled
control subjects. This is in agreement with previous
reports on the familiality of AS, but the present study has
more power and extends over larger familiar cohorts than
previously reported.
INTRODUCTION
Ankylosing spondylitis (AS) is a chronic infl am-
matory disease, characterised by low back and
buttock pain with morning stiffness of insidious
onset.1 Population-based studies suggest that the
prevalence of AS is between 0.1% and 1.4%1–3 and
we have found the prevalence of AS in Iceland to
be close to 0.13%.4 Clinicians are well aware of
the increased prevalence of AS among siblings of
patients with AS, an observation confi rmed in a
number of studies.5–7
Several studies demonstrate that 6% to 8% of
close relatives of patients with AS have the dis-
ease5–7 in comparison to 0.1% to 1.4% in the gen-
eral population.1–4 The prevalence of AS among
relatives to human leucocyte antigen (HLA)-B27
positive patients with AS is even higher or up to
20%.8 Studies of twins have shown the concor-
dance rate in monozygotic twins to be 40% to
75%,9–11 but 4.3% to 12.5% in dizygotic twins,11
refl ecting that genetic and environmental factors
are important in the pathogenesis of AS. Only two
reports have reported the prevalence of AS among
second-degree relatives (SDRs) and third-degree rel-
atives (TDRs) of patients with AS, respectively.12 13
Several of these studies have had various method-
ological limitations, for example, limited cohort
size, index cases frequently selected from specialty
A strong familiality of ankylosing spondylitis through
several generations
Arni Jon Geirsson,1 Kristleifur Kristjansson,2 Bjorn Gudbjornsson3,4
1Department of Rheumatology,
University Hospital, Reyjavík,
Iceland
2deCODE Genetics, Reykjavik,
Iceland
3Centre for Rheumatology
Research, University Hospital,
Reyjavík, Iceland
4University of Iceland, Reykjavik,
Iceland
Correspondence to
Bjorn Gudbjornsson, Centre
for Rheumatology Research,
Landspitali, Hringbraut –
University Hospital,
Reykjavik 101, Iceland;
bjorngu@landspitali.is
Accepted 6 February 2010
clinics, lack of well documented prevalence of AS
in the background population and the ascertain-
ment bias of looking for fi rst-degree relative (FDR)
cases within a family after identifying the index
case, which can result in infl ated λS values, as sug-
gested by Guo.14
To overcome some of these methodological
problems, we have recently identifi ed all known
cases with AS in Iceland4 and by using the Icelandic
genealogy database of deCODE Genetics (Reyk-
javík, Iceland), we are able to calculate the RR for
family members spanning several generations of
individuals with AS.
MATERIALS AND METHODS
The study group
The study involved all known patients with AS in
Iceland in 2005. Patients were recruited from three
main sources. The fi rst source was from a database
of individuals participating in genetic studies of
AS and infl ammatory bowel diseases.13 From this
database, 205 individuals were included in the pre-
sent study. The second source was an electronic
registry of patients admitted to the two major hos-
pitals in Iceland (WHO International Classifi cation
of Diseases, 10th edition (ICD-10) codes: M 45,
M 45.5, M45.9, M 46 and M 46.9), Landspitali –
University Hospital in Reykjavík and the University
Hospital in Akureyri. This source yielded an addi-
tional 54 patients with AS. The third source was a
personal call to all private outpatient rheumatology
services in Iceland to report patients to the study,
yielding 64 additional patients with AS.
The above-mentioned 3 sources yielded in total
280 individuals with AS, of which 256 were alive in
December 2005. A detailed description of the inclu-
sion criteria for verifi ed AS according to the modi-
fi ed New York criteria for classifi cation of AS15 can
be found in our previous work on the prevalence of
AS in Iceland.4
Genealogy database
deCODE Genetics has built a computerised
genealogy database of more than 760 000 indivi-
duals.16 The database contains records of all living
Icelanders, comprising more than 300 000 individ-
uals and a large proportion of all individuals who
have ever lived in the country from the time of the
settlement in the late ninth century. The genealogy
database is essentially complete from the 18th cen-
tury to the present day, allowing distant relation-
ships to be traced accurately.
Use of this database allows the defi nition of all
the relationships between patients with AS, as well
as the degree of the relatedness. It also allows the
19_annrheumdis125914.indd 134619_annrheumdis125914.indd 1346 6/9/2010 11:50:56 AM6/9/2010 11:50:56 AM
group.bmj.com on June 21, 2010 - Published by ard.bmj.comDownloaded from

Concise report
Ann Rheum Dis 2010;69:1346–1348. doi:10.1136/ard.2009.125914 1347
creation of matched control groups for calculation of the relative
risks and their statistical signifi cance. To ensure anonymity of
the patients in the present study, the social security numbers of
participants were encrypted by the Data Protection Commission
of Iceland before being used in the analyses.17 The study proto-
col was approved by the Icelandic Bioethics Committee and the
Icelandic Data Protection Commission.
Assessment of inheritance
The RR for disease in relatives is a measure of the risk of
disease in a relative of an affected person as compared with
the risk in the population as a whole. Obtaining valid esti-
mates of the RR is, however, not straightforward because
many sampling schemes may lead to biased or inaccurate
estimates.14 The use of a population-based group of patients
eliminates some of this potential sampling bias. In addition,
a near-complete genealogy database facilitates identification
of virtually all relationships between patients. It is impor-
tant to note that only probands were used in our analysis,
and no attempt was made to recruit relatives of cases of AS.
This design avoids the potential overestimation of familiar-
ity when secondary cases are recruited through probands, as
described by Guo.14
In the present study, we determined the RR for AS in a previ-
ously described Icelandic population.4 To assess the signifi cance
of the RR obtained for a given group of patients, we compared
their observed values with the RR computed for up to 1000
independently drawn and matched groups of control indivi-
duals.18 Each patient was matched to a single control individ-
ual in each control group. The controls were drawn at random
from the genealogy database and matched for the year of birth,
gender and the number of ancestors recorded in the genealogy
database. Empirical p values can be calculated using the con-
trol groups; thus, a p value of 0.05 for the RR would indicate
that 5% of the matched control groups had values as large as
or larger than that for the patient’s relatives or spouses. The
number of control groups required to obtain a fi xed accuracy of
the empirical p values is inversely proportional to the p value.
We therefore selected the number of control groups generated
adaptively up to a maximum of 1000. When none of the values
computed for the maximum number of control groups were
larger than the observed value for the patient’s relatives and
spouses, we report the p value as being less than 0.001. Using a
variance stabilising square root transform, an approximate 95%
CI may be constructed based on the distribution of RR for con-
trol groups.
RESULTS
Pedigrees
Figure 1 shows an example of an ancestral pedigree with 18
patients with AS (in black) spanning 10 generations. In this case,
102 patients were related at or within the distance of only 4
meioses in a total of 45 families. The pedigrees demonstrated
signifi cant clustering of patients with AS. Additional pedigrees
can be obtained from the authors on request.
RR in relatives
The RR estimates for disease in relatives of affected patients
with AS are shown in table 1. The RR for AS were 75.5, 20.2
and 3.5 (all p values<0.0001) among FDRs, SDRs and TDRs of
affected patients, respectively. The RR for the fourth-degree
relatives did not reach signifi cance levels: RR 1.04 (95% CI 0.58
to 1.66), p value=0.476.
If the degree of relatedness includes aunts and uncles, nephews
and nieces, or cousins, the RR is similar to the above- presented
data (further details are presented in table 2).
Sex differences
Calculation of the RR for AS in siblings of affected parents with
AS in relation to whether the they were paternal or maternal
index cases demonstrated a similar RR; paternal siblings had a
RR of 48.2 (95% CI 36.22 to 63.19) and maternal siblings a RR
of 43.1 (95% CI 31.52 to 57.24). In contrast, the father of a case
with AS had a much higher RR than did the mother of the index
case: 167 (95% CI 131 to 194) versus 48.9 (95% CI 24.31 to
63.90).
Table 1 Relative risk* estimates of ankylosing spondylitis (AS) in
relatives in four generations of affected individuals with verifi ed AS
Degree of relationship RR (95% CI) p Value†
Number of
affected
relatives
First-degree relatives 75.49 (60.19 to 93.87) <0.001 112
Second-degree relatives 20.21 (15.15 to 30.20) <0.001 50
Third-degree relatives 3.52 (2.20 to 5.67) <0.001 28
Fourth-degree relatives 1.04 (0.58 to 1.66) <0.476 18
*Risk estimates for AS are the RR (95% CI) for the estimated risk of having the disease
itself (eg, the risk of AS in a relative of a patient with AS) as compared with the risk in
1000 sets of Icelandic control subjects.
†p Values are one sided and indicate the signifi cance of the risk of disease in the
combined relative groupings of all generational levels as compared with matched
control subjects.
Table 2 Relative risk* for having ankylosing spondylitis (AS) for aunts
and uncles, nephews and nieces, or cousins, of affected individuals with
verifi ed AS
Relationship RR (95% CI) p Value†
Number of
affected
relatives
Aunts and uncles 18.29 (12.93 to 27.63) <0.001 20
Nephews and nieces 18.12 (12.41 to 27.63) <0.001 20
Cousins 3.66 (2.27 to 5.72) <0.001 26
*Risk estimates for AS are the RR (95% CI) for the estimated risk of having the disease
itself (eg, the risk of AS in a relative of a patient with AS) as compared with the risk in
1000 sets of Icelandic control subjects.
†p Values are one sided and indicate the signifi cance of the risk of disease in the
combined relative groupings of all generational levels as compared with matched
control subjects.
Figure 1 Pedigree of patients with ankylosing spondylitis (AS)
extending over 10 generations. Disease status is known only for the
later generations. This pedigree was created with the use of the deCODE
genetics genealogy database. To protect the anonymity of the families,
some of the unaffected relatives in the pedigree are not shown. The
circles denote female family members and the squares denote male
family members.
19_annrheumdis125914.indd 134719_annrheumdis125914.indd 1347 6/9/2010 11:50:56 AM6/9/2010 11:50:56 AM
group.bmj.com on June 21, 2010 - Published by ard.bmj.comDownloaded from

Concise report
Ann Rheum Dis 2010;69:1346–1348. doi:10.1136/ard.2009.1259141348
RR in spouses
The RR of spouses of patients was 6.64 (95% CI 0.88 to 12.80),
but this value did not reach signifi cance (p value=0.097). The RR
for spouses of the controls was 0.952.
DISCUSSION
In the present study, we have used an extensive genealogical
database, extending back to the late ninth century, for evaluat-
ing the relationships of all known patients with AS in Iceland.
The study demonstrates a strong familiality of AS over three
generations, but the risk of developing AS seems to disappear in
the fourth generation of individuals with AS. The present results
are the fi rst to illustrate signifi cant familiality of AS over sev-
eral generations in a population-based cohort. Our study con-
fi rms results of previous studies of familiality of AS that have
shown an increased prevalence of AS among siblings and FDRs
of patients with AS.5–7 12 13 These studies have illustrated a high
RR of AS in siblings and FDRs of individuals with AS, with a λS
(sibling recurrence RR=λS) of up to 82 and λR2 (SDRs recurrence
RR) of 10, compared to 75 and 20, respectively, in our study.
To our knowledge, only two studies have reported the prev-
alence of AS in SDRs and TDRs of individuals with AS.5 13 In
1961, de Blecourt and coworkers published in the Annals of
Rheumatic Diseases the prevalence of AS among 2862 relatives of
100 patients with AS and reported an 86% inclusion rate among
relatives. They found that the RRs were around 10 for SDRs and
7.0 for TDRs, while we found a RR of 20 and 3.5 in SDRs and
TDRs, respectively. In our study, we identifi ed all known AS
cases in Iceland and therefore were neither dependent on clinical
evaluation of all the relatives of our patients nor on the preva-
lence value of AS in Iceland, which may explain the differences
in our results. We report the RR for FDRs, SDRs and TDRs and
also report that the risk of having AS mostly disappears in the
fourth generations of our 259 index cases (RR 1.04; p=0.476),
which is lower compared to previous report from Iceland on
this issue.13
The observation of a RR of 75 and 20 for FDRs and SDRs,
respectively, to develop AS is far higher than has been observed
using the same methods for rheumatoid arthritis in Iceland, or
4.38 for FDRs and 1.95 for SDRs,19 respectively and also well
higher than we have reported for psoriatic arthritis in Iceland
or 39.2 for FDRs and 12.2 for SDRs.20 Thus, the present results
demonstrate that genetic factors in AS probably play a stron-
ger role in the development of AS than in rheumatoid arthritis
or psoriatic arthritis, but that the genetic effects in all of these
diseases extend over several generations. However, our study
was not designed to elucidate any collective environmental or
immunological factors, for example, HLA-B27, that may play a
role in the pathogenetic processes of AS.
The prevalence of AS is much higher among men than women
(4 to 1). Therefore, we looked at whether the RR of siblings was
dependent on whether their mother or their father had AS. No
signifi cant difference in their RR was found in this respect.
All previous studies on familiality of AS have either presented
only the prevalence value of AS among relatives to patients with
AS or relied on estimates of the population prevalence of AS
when calculating the λ values. In contrast, we based our study
on complete ascertainment of all known cases of AS in Iceland as
we did in our recent population-based study on the prevalence
of AS in Iceland.4 This we believe may alleviate some of the
ascertainment bias that is inherent in studies that collect index
cases and affected relatives for estimation of the familiar-related
RR from specialised outpatient clinics only. These studies are
likely to over-represent patients with more severe disease and
families with more cases, which will infl uence the degree of her-
itability.14 Meanwhile, our study used an unselected nationwide
group of all known patients with verifi ed AS in Iceland.
In conclusion, individuals with AS in Iceland are signifi cantly
more related to each other than randomly sampled Icelandic
subjects for three generations. The present study used large,
unselected family cohorts combined with an extended geneal-
ogy database to avoid the bias of traditional post hoc ascertain-
ment of family members. These fi ndings suggest that genetic
factors play a stronger role in AS in comparison to rheumatoid
arthritis and psoriatic arthritis.
Acknowledgements The authors would like to thank the staff of the Statistics
Department at deCODE Genetics for their assistance with statistical analyses, espe-
cially Valdimar Búi Hauksson.
Funding The study was funded by the research foundations of the University
Hospital in Iceland, the Society for Rheumatology in Iceland and the Wyeth
Rheumatology Foundation in Iceland.
Patient consent Obtained.
Ethics approval This study was conducted with the approval of the Icelandic
Bioethics Committee and the Icelandic Data Protection Commission.
Provenance and peer review Not commissioned; externally peer reviewed.
REFERENCES
1. Gran JT, Husby G. The epidemiology of ankylosing spondylitis. Semin Arthritis Rheum
1993;22:319–34.
2. De Angelis R, Salaffi F, Grassi W. Prevalence of spondyloarthropathies in an
Italian population sample: a regional community-based study. Scand J Rheumatol
2007;36:14–21.
3. Gofton JP, Robinson HS, Trueman GE. Ankylosing spondylitis in a Canadian Indian
population. Ann Rheum Dis 1966;25:525–7.
4. Geirsson AJ, Eyjolfsdottir H, Bjornsdottir G, et al. Prevalence and clinical
characteristics of ankylosing spondylitis in Iceland – a nationwide study. Clin Exp
Rheumatol 2010. In press.
5. de Blecourt J, Polman A, de Blecourt-Meindersma. Hereditary factors in rheumatoid
arthritis and ankylosing spondylitis. Ann Rheum Dis 1961;20:215–20.
6. LeClercq SA, Chaput L, Russell AS. Ankylosing spondylitis: a family study.
J Rheumatol 1983;10:629–32.
7. Bedendo A, Glorioso S, Venturi Pasini C, et al. A family study of ankylosing
spondylitis. Rheumatol Int 1984;5:29–32.
8. van der Linden SM, Valkenburg HA, de Jongh BM, et al. The risk of developing
ankylosing spondylitis in HLA-B27 positive individuals. A comparison of relatives of
spondylitis patients with the general population. Arthritis Rheum 1984;27:241–9.
9. Järvinen P. Occurrence of ankylosing spondylitis in a nationwide series of twins.
Arthritis Rheum 1995;38:381–3.
10. Brown MA, Kennedy LG, MacGregor AJ, et al. Susceptibility to ankylosing
spondylitis in twins: the role of genes, HLA, and the environment. Arthritis Rheum
1997;40:1823–8.
11. Pedersen OB, Svendsen AJ, Ejstrup L, et al. Ankylosing spondylitis in Danish
and Norwegian twins: occurrence and the relative importance of genetic vs.
environmental effectors in disease causation. Scand J Rheumatol 2008;37:120–6.
12. Brown MA, Laval SH, Brophy S, et al. Recurrence risk modelling of the genetic
susceptibility to ankylosing spondylitis. Ann Rheum Dis 2000;59:883–6.
13. Thjodleifsson B, Geirsson AJ, Björnsson S, et al. A common genetic background for
infl ammatory bowel disease and ankylosing spondylitis: a genealogic study in Iceland.
Arthritis Rheum 2007;56:2633–9.
14. Guo SW. Infl ation of sibling recurrence-risk ratio, due to ascertainment bias and/or
overreporting. Am J Hum Genet 1998;63:252–8.
15. Brewerton DA, Hart FD, Nicholls A, et al. Ankylosing spondylitis and HL-A 27. Lancet
1973;1:904–7.
16. Gulcher J, Kong A, Stefansson K. The genealogic approach to human genetics of
disease. Cancer J 2001;7:61–8.
17. Gulcher JR, Kristjánsson K, Gudbjartsson H, et al. Protection of privacy by third-party
encryption in genetic research in Iceland. Eur J Hum Genet 2000;8:739–42.
18. Reynisdottir I, Gudbjartsson DF, Johannsson JH, et al. A genetic contribution to
infl ammatory bowel disease in Iceland: a genealogic approach. Clin Gastroenterol
Hepatol 2004;2:806–12.
19. Grant SF, Thorleifsson G, Frigge ML, et al. The inheritance of rheumatoid arthritis in
Iceland. Arthritis Rheum 2001;44:2247–54.
20. Karason K, Love TJ, Gudbjornsson B. A strong heritability of psoriatic arthritis
over four generations. The Reykjavik psoriatic arthritis study. Rheumatology
2009;8:1424–8.
19_annrheumdis125914.indd 134819_annrheumdis125914.indd 1348 6/9/2010 11:50:57 AM6/9/2010 11:50:57 AM
group.bmj.com on June 21, 2010 - Published by ard.bmj.comDownloaded from