Prevalence of Spondylarthropathies

7/29/2019 Prevalence of Spondylarthropathies

http://slidepdf.com/reader/full/prevalence-of-spondylarthropathies 1/10

58

ARTHRlTI5 & RHI UMATISMVol 41. N o 1, January 1998, pp 58-670 998, AineriLan College of Rheumdtology

PREVALENCE OF SPONDYLARTHROPATHIES INHLA-B27 POSITIVE AND NEGATIVE BLOOD DONORS

JURGEN BRAUN, MATHIAS BOLLOW, GEROLD REMLINGER, ULRTCH EGGENS,MARTIN RUDWALEIT, ARMTN DISTLER, and JOACHIM SIEPER

Objective. To determine the overall prevalence of

spondylarthropathy (SpA) among whites.

Methods. To screen for SpA symptoms, such as

inflammatory back pain (IBP), joint swelling, psoriasis,and uveitis, or a specific family history, questionnaires

were mailed to 348 blood donors (174 HLA-B27 positive

and 174 HLA-B27 negative). From the responding 273

persons (78%; 140 B27 positive, 133 B27 negative), 126

were selected for further evaluation based on the symp-

toms reported. Of this group, 90 persons agreed to

undergo physical examination (71.4%; 46 B27 positive,

44 B27 negative). There was no difference between the

B27-positive and -negative groups in terms of age (mean

f SD 38.4 f 10 versus 39.5 & 11years) and sex ratio

(67% versus 68% were men). In addition, 58 donors (32

B27 positive, 26 B27 negative) agreed to undergo mag-netic resonance imaging (MRI) of the sacroiliac joints.

A diagnosis of SpA and ankylosing spondylitis (AS) was

made according to the European Spondylarthropathy

Study Group criteria and the New York criteria.

Results. SpA was diagnosed in 20 persons: 19 of

140 B27-positive (13.6%) and 1 of 133 B27-negative

(0.7%) subjects (15 male and 5 female). AS was diag-

nosed in 9 persons (7 male and 2 female; 45%), undif-

ferentiated SPA (USpA) in 7 (5male and 2 female; 35%),

psoriatic arthritis (PsA) in 3 (2 male and 1 female;

15%), and chronic reactive arthritis (Re& Reiter’s

Drs. Braun and Slepcr’s work was supported by grants fromthe E ckehard B ode Stiftung, the Berliner Sparkassenstiftung Medizin,and from the Bundcsminister ium fur Forschung und Technologie(BMBF).

Jurgen Braun, M D, G erold Remlinger , Ulrich Eggens, MartinRudwaleit, Arrnin Distler , MD, Joachim Siepcr , MD: Klinikum Bcn-jamin Franklin Free University, and Deutschcs Rheumalorschung-szentrurn, Berlin, Germany; Mathias Bollow, MD: Klinikum Charite,Hurnbo ldt University, Berlin, Germany.

Address reprint requests to J iirgcn Braun, MD , Rheumatolo-gie, Abteilung fur Allgemeinc lnnere Mcdizin und Ncphrologie,Universititsklinikum Benjamin Frank lin Frc c University, Hinden burg-damm 30, 12200 Berlin, Germany.

Subm itted for publication May 19, 1997; accepted in revisedlorm A ugust 6, 1997.

~ .~

syndrome) in 1 (male; 5%). On the basis of a B27

frequency of 9.3% among the population of Berlin (3.47

million persons), the estimated prevalence of SpA was

1.9%, AS was 0.86%, USpA was 0.67%, and PsA was0.29%. The relative risk of developing SpA in B27-

positive subjects was calculated as 20.7 (95% confidence

interval 4.6-94.2; P =0.001). Of 58 persons with IBP,

sacroiliitis was detected by MRI in 15 of 32 B27-positive

(46.9%) and 1 of 26 B27-negative (3.9%) subjects (P=

0.002). Four of these 16 donors did not fulfill diagnostic

criteria for SPA.

Conclusion. With a calculated prevalence of 1 9 0 ,

spondylarthropathies are among the most frequent

rheumatic diseases in the white population. HLA-B27

positive persons carry a 20-fold increased risk of devel-

oping SPA. AS and USpA are the most frequent SPAsubtypes. Persons with IBP who are B27 positive have a

50% likelihood of having sacroiliitis.

The term “~pondylarthropathy~~SpA) covers agroup of inflammatory rheumatic diseases (1) compris-

ing ankylosing spondylitis (AS), the prototype of the

spondylarthropathies, reactive arthritis (ReA), psoriatic

arthritis (PsA), arthritis associated with inflammatory

bowel disease (AIBD), and undifferentiated SPA(USpA), an only recently recognized subset (2). The

spondylarthropathies share typical clinical featurcs, such

as inflammatory back pain (IBP) ( 3 ) ,which in the earlystages often alternates in the buttocks due to inflamma-

tion of the sacroiliac (SI) joints ( 44, eripheral oligo-arthritis predominantly of the lower extremities, enthe-

sopathy, anterior uveitis, and a striking association withHLA-B27 (6-8). How this major histocompatibility

complex (MHC) class I molecule contributes to thepathogenesis of SPA is not known (9). The magnitude of

the HLA-B27 association differs between SPA subsets,and it correlates with the presence of certain clinical

features, such as involvement of the axial skeleton, andwith disease severity ( 5 ) .



HLA-B27 AND SPONDYLARTHROPATHIES 59

There has been a lot of controversy about the

clinical importance of HLA-B27 determinations, and its

value is still questioned, not only because of the rela-

tively high costs of the procedure (10). While some

auth ors deny almost any clinical value (3,11),others find

it useful to know the B27 status in doubtful diagnostic

situations (12J3). It is clear that determination of B27

subtypes is o f no clinical importance (14).

The overall prevalence of SPA among whites is

not known. Data are only available f r o m Eskimo popu-

lations (15,16) and suggest a prevalence of -2%. How-

ever, the B27 frequency of >20% in these populations

(17,18) is different f r o m the f requency in white

populations.

Epidemiologic surveys have been performed be-

fo re (19,20) and afte r (21-25) repor ts of B27 association

with AS (6) and Reiter’s syndrome (7) to estimate theprevalence of SPA. The best-studied SPA subset over

the last 3 decades has been AS. T he prevalence data f o r

AS have shown so me variance, partly du c to differences

in the study design. Studies of hospital records, blood

donors, or population prevalence have yielded rates that

vary between 0.1% in Holland (23) and 1.1-1.4% in

Norway (24). Even higher rates have been reported in

American blood donors, suggesting that 18% (21) and

more (22) of the B27-positive blood donors examined

had AS. These latter studie4 have been intensively

critiqued (26,27).

Similar to Calin and Fries’ study conducted 20years ago (21), for the present study, we chose apopulation-based approach and evaluated “healthy”

blood donors with known H LA status, mailing question-

naires to potential study subjects. However, in contrast

to C alin’s and m ost oth er published studies dealing with

the prevalence of AS, we screened f o r the wider spec-

trum of SPAby using the Eur op ean Spondylarthropathy

Study Group (ESSG) criteria. In addition to a medical

history and physical examination, the techn ique of m ag-

netic resonance imaging (MRI) of the SI joints recently

developed by our group (28-30) was uscd to diagnose

sacroiliitis in HLA-B27 positive and negative personswho reported IBP.

SUBJECTS AND METHODS

Blood donors were recrui ted fro m 5 blood banks: theuniversity hospitals Benjamin Franklin and Free University;Rudolf Virchow and CharitC, both at Humboldt University;and the hospitals in Berlin-Friedrichshain and Berlin-Buch.Fr om a total of 1,871 registered blood d onor s, 174 B27-positivepersons were identified, and the addresses of these d onors andof 174 age- and sex-matched B27-negative blood d ono rs were

obtained. Thirteen donors in the B27-positive and 1.5 in the

B27-negative group could not be reached by mail or phonebecause they had relocated and had left no forwarding address.

Thus, questionnaires were mailed to 161 B27-positive and 159

B27-negative subjects (n 320).The questionnaire covered subjective symptoms ofback pain, joint pain and swelling, peripheral arthritis, uveitis,psoriasis, a nd frequency of diar rhe a and dysuria. All questionsnecessary to diagnose IBP were asked (initial age at the start ofthe first symptoms, mode of onset, duration of back pain,mornin g s ti ffness , whethe r thcr c was im provem ent withexerciseh-est). A family history of rheumatic disease, especiallyAS, psoriasis, chronic IB D, an d uveitis, was assessed. Fur therquest ions concerned preceding trauma and intensi ty and du-ration of pain.

The responses were evaluated without knowledge ofthe subject’s HL A status, and 12 6 persons with any 1 symptomsuggestive of SPA (TBP, perip hera l ar thritis of th e lower limbs,

enthesopathy, dactylitis, uveitis, psoriasis, diarrhea > time/month, urethritis >1 time/year, family history of rheumaticdisease) were identified. These wc rc all contacted by tele-pho ne, if available, or by mail and asked to a tten d the hospitalfor physical examination. This examination comprised mea-sure men t of anterior spinal flexion according to Schohe r (31)

and measurement of thoracic expansion (32).All persons were initially seen by a specially trained

medical s tudent (GR ), and if there were serious complaints ora diagnosis of S pA was suspected, a rheumatologist saw thcsubject (JB o r UE). Psoriatic lesions were either documentedby one of us or had previously been diagnosed by a dermatol-ogist. Similarly, uveitis was considered to be present only ifther e was an ophthalmologic diagnosis of anterior uveitis. Th edon ors were also asked ab out the num ber of days missed from

work the previous year because of their symptoms.Conventional radiographs of the pelvis and the spine

that had been obtained not more than 2 years previously, ifavailable, were reevaluated. Established grading systems (33)

were used for the independent evaluations by a radiologist

(MB ) and a rheumatologist (JB or U E) who were not aw are ofth e clinical €indings.

Pers ons with an swers suggestive of TBP (3) were askedto undergo dynamic MRI of the SI joints. The M R I wasperformed as recently described (28), and the evaluation wasdone as recently proposed (30). For an M RI diagnosis ofdefinite sacroiliitis, ei ther ac ute changes with an enha nce me ntof >40% or chronic changes >grade 11(30) wcre necessary.The contrast agent Gadolinium DTPA was kindly provided by

Pharm a Deutchland (Schering AG , Berl in, Germany).A diagnosis of SpA was made if th e ESS G criteria were

fulfilled (1). USpA was diagnosed if the ESSG criteria werefulfilled, but no diagnosis of AS, PsA, ReA. or AIB D could bemade. PsA was diagnosed if the E SSG criteria were fulfilled,and typical skin or nail lesions were present.

A diagnosis of AS was mad e if th e modified New Y orkcriteria were met: radiog raphs showing bilateral changcs in theSI jo ints >grad e I1 yielded a diagnosis of definite AS (33). Th eAmerican College of Rheumatology 1981 criteria for Reiter’ssyndrome (34) were fulfilled by 1 male donor with chronicurethritis and interm ittent pain and swelling of the knee joint,wh o was classified as having chron ic Re A.



60 B R A U N ET AL

Table 1. Overvicw of the study”

No . (76) of No. (56) No . (%)

patients HLA-927- HLA-B27-

Addresses identified 348 174 174Questionnaires sent 320 (92.0) 161 (92.5) 159 (91.4)Responders 273 (85.3) 140 (87.0) 133 (83.6)Suspected of having Sp A 126 (46.2) 6X (48.6) 58 (43.6)Physical cxamination 90 (71.4) 46 (67.6) 44 (75.9)MRI examination 58 (64.4) 32 (6Y.6) 26 (59.1)

‘*SpA =spondylarthropathy; MRT =magnetic resonaiice imaging.

RESULTS

Study population. The overall response rate to

the initial questionnaire was 85.3% (273 of 320), 140

subjects in t he B 27-positive grou p (87% ), and 133 in the

B27-negative group (83.6%) (P not significant). Therc

was no significant difference between these groups interms of age and sex. Th e mean ? SD age was 38.7 ? 11

and 38.3 2 11 years, respectively, and the percentage of

male subjects was 67.9% and 66.296, respectively.

Symptoms suggestive of SPA w ere found in 126

of the 273 subjects (46.2%), 66 were B27 positive and 60

were B27 negative. Of these 126 subjects, we were

unable to contact 11 for various reasons and another 25

declined to attend the hospital (20 B27-positive, 16

B27-negative subjects). Thus, 90 of the 126 blood d onors

with possible SP A (71.4%) underwen t physical cxamina-

tion. Ninety-four of the 126 subjects with possible SPA

(74.6%; 54 B27 positive, 40 B27 negative) reportedsymptoms suggestive of IBP. Of these 94 subjects, 76

came for physical examination, an d 58 of these subjects

agreed to undergo MRI. Thus, 32 of these 126 blood

donors had signs othe r than IR P that were suggcstive of

SPA. An overview is given in Table 1.

Identification of patients with SPA. On the basis

of the history, physical examination, and pelvic radio-

graphs, SPA was diagnosed in 20 of the 273 question-

naire responders (7.3%), 19 of 140 (13.6%) in the

B27-positive group and 1 of 133 (0.7%) in the B27-

negative group (15 male and 5 female; ma1e:female ratio

3:1) (Figure 1). With respect to the SP A subsets, AS wasdiagnosed in 9 of the 140 B27-positive donors (7 male

an d 2 female; 6.4%) and U SpA in 7 (5 male and 2 female;

5 % ) . PsA was diagnosed in 2 male (1 B27 negative) an d

1 female (B27 positive) subject, and chronic ReA (Re-

iter’s syndrome) was diagnosed in 1 male subject (B27

positive). An overview is given in Table 2.

Definite radiographic changes according to theNew Yo rk criteria were seen in all 9 subjects with AS. In

6 of these subjects, the diagnosis of AS was previouslyknown. In the remaining 3 donors, who had severe

*O51 pto.001nHLA B27+ HLA 8127-(n=140) (n=133)

Figure 1. Spondylarthropathy in 273 blood donors. The percentages

of subjects with a diagnosis of spondylarthropathy (as defined in

Subjects and Mcthods) who were HLA-B27 positive and HLA-B27

negative are shown. The difference was statistically significant.

chronic changes on MRI (see below), radiographs were

performed for comparison and confirmed the MRI

results. Of the 7 donors diagnosed as having USpA, 3

had been told that they had “some recurrent, possibly

chronic, form of arthritis.” Of the 3 donors with PsA, 2

were aware of the diagnosis. In the donor with Reiter’ssyndrome, the diagnosis was suspected. Overall, acute

and/or chronic changes in the SI joints were detected by

MRI in 10 of the 16 examined subjects diagnosed as

having SP A. Th e characteristics, sym ptoms, and findings

in the persons diagnosed as having SPA are shown in

Table 3.

MRI of the S I joints. A total of 58 blood donors

were evaluated by M RI , 32 were B27 positive (25 male,

7 female) and 26 were B27 negative (21 male, 5 female).

Table 2. Diagnoses and calculated prevalence rates*

No . of Prevalence (%) Calculated prevalence (5%)

patients in B27+ patients in the population of Berlin

SPA 20 13.6AS 9 6.4USpA 7 5.0

PsA 3 I .4Re A 1 0.7

1.90.860.670.290.1

”The prevalcnce of HLA-B27 was based on the 140 B27+ blooddonors examined. The calculated prevalence was determined asdescribed in the Results section. SpA = spondylarthropathy; AS =

ankylosing spondylitis; USpA =undifferentiated spondylarthropathy;PsA = psoriatic arthritis; ReA =reactive arthritis.




Table 3. Characteristics of the blood donors diagnosed as having spondylathropahty (SPA)*

M R I

chronicitylactivity

index?

Duration ofPatient1 symptoms Inflammatory Peripheral Family Right SI Left SI

agelsex Diagnosis (years) HLA-B 27 back pain arthritis SpA features history joint joint

13138lM14133lF

251231M33143F

38135IM

41149lM641331M891371M701421M241561M42155lM621251M83147lF

921431M54158lM781341F821331M

771441M941331M5214.51F

ASASASASASASASASASUSpAUSpAUSpAUSpAUSpAUSpAUSpAChronic

R eAPs APsAPs A

84

13

13

1424128

173.5388

30

20259

13

253

30

+++++++++++++

++++

+

+-

+++++++++++++

++++

+++

---Knee----Ankle, kneeArthralgia---ToeKnee---KneeHand, feet

- - III/X 1n;xDiarrhea - IIIA 11B- - IIIB IIIB

Psoriasis ~ IIIIX 1rr/x- - IIIIA IIIIAUvcitis - IIIX IIiX- - IlIlB IIIIB- - IVIX lVlB

Alternating buttock pain - IIX IIIX

Uveitis AS IIA IIIAUveitis, alternating buttock pain - IIIA OIX- Ps OIX IIB

Dactylitis - IIIB OIXUveitis, diarrhea - ND

Uveitis, diarrhea, urethritis - OIB OIB

Uvcitis, psoriasis - 01x OIX- l/X

- - N D

Enthesopathy Ps N D

IiX DPsoriasis, diarrheaPsoriasis Ps

*AS = ankylosing spondylitis; USpA = undifferentiated spondylarthropathy; Ps = psoriasis; PsA =psoriatic arthritis; ReA = reactive arthritis.t The magnetic resonance imaging (MRT) grading system is explained in detail in ref. 30. Short form grading: 0 = normal sacroiliac (SI) joints; I

= minor changes; I1 = small erosions, subchondral sclerosis, fat accumulations; 111 = severc erosions; and IV = ankylosis. Grading aftergadolinium-DTPA ad ministration: X = no enhancement, A =moderate enhancement, and B =strong enhancement. ND = not done.

All had complaints of back pain of the IBP type.

Evidence of past or present sacroiliitis was obtained in

15 B27-positive subjects (46.9%) and 1 B27-negative

subjects (3.9%) ( P =0.002) (Figure 2). For comparison,

signs of degenerative spinal disease were detected at a

similar frequency in both groups (15.6% versus 19.2%).

Figures 3 and 4 are examples of MRTs, on e of a subject

with AS and the other of a subject with U SpA.

Amo ng the 16 donors with MRI-proven sacroili-

itis, 12 fulfilled th e diagnostic criteria for SPA, and 4 did

not (3 B27 positive an d 1 B27 negative).

Risk of SPA in B27-positive subjects. The odds

ratio of having past or present SPA in B27-positive

subjects compared with B27-negative subjects was 20.7

(95% confidence interval 4.6-94.2; P = 0.001).

Frequency of B27 in the population of Berlin.

Amo ng th e 1,871 blood do nors registered in Berlin, 174

were fo und to be B27 positive. This suggests an overall

B27 prevalence of 9.3% in the general population of

Berlin. The relative frequency of B27 was similar in the

5 blood banks part icipating in the s tudy ( range

8.9-9.5%).

%

pt0.0011 m

HLA B27+ HLA 627-

(n=32) (n=26)

Figure 2. Detection of sacroiliitis by magnetic resonance imaging

(MRI) in 58 blood donors who agreed to undergo imaging. The

frequen cies of detection of sacroiliitis by MR I in HLA-B27 positive

and HLA-B27 negative blood donors a re shown. The difference was

statistically significant.



62 BRAUN ET AL

Figure 3. Magnetic resonance imaging of a 35-year-old male blood

donor with bilateral sacroiliitis grade IIIB (patient 38 in Table 3),

diagnosed as having ankylosing spondylitis. T1-weighted image (500/20

ms, 90“). Thcre are juxtaarticular erosions (arrows), irregular joint

spaces with hypointense areas of subchondral inflammation (arrow-

heads) . and fat accumulations (asterisks) indicating chronicity. The

dynamic sequence (not shown) revealed a high level of enhancement

after administration of the contrast agent.

Calculation o f SpA prevalence in the p opulation

o f Berlin. O n Dece mber 31, 1995, the total population of

Berlin was 3,471,418, with 1 8% < 18 y ears old an d 13.8%

>65 years old (official data from the “StatistischesLandesamt Berlin,” Kulturbuch Verlag Berlin). On the

basis of a B27 frequency of 9.3% among the 2,367,507

adult (ages 18-65) residents of Berlin, the overall prev-alence of SPA wa, estimated to b e 1 .9%. Thus, out of

220,178 B27-positive and 2,147,329 B27-negative Be rlin-ers between 18 and 65 years of age, 44,975 are expec ted

to have SPA (29,944 of the B27-positive and 15,031 ofthe B27-negative residents).

If we had overestimated the SPA prevalence inB27-negative subjects 2-fold (and therefore only 7,765

B27-negative Berliners had SpA), the overall prevalenceof SPAwould still be 1.6%.

If the 3 B27-positive donors with MRI-proven

sacroiliitis were diagno sed as having SPA, the prev alenceof SPA amon g B27-positive subjects would be 15.7% (22of 140 HLA-B 27 positive subjects). This would result inan overall SpA prevalence of 2.1%.

Figure 4. Magnetic resonance imaging of a 43-year-old male blood

donor with unilateral sacroiliitis grade IIA (patient 92 in Table 3),

diagnosed as having undifferentiated spondylarthropathy (same T1-

weighted sequence as in Figure 3). The left sacroiliac joint is normal.

The right joint shows an irregular joint space and erosions (arrows),

with the fat accumulation (asterisk) indicating a chronic change. The

intcrvertebral disc space bctween L5 and S l is narrow and shows a line

of low signal intensity (arrowhead). Such vacuum phenomena are seen

in degenerative spinal diseases (spondyloosteochondrosis). The dy-

namic sequencc (not shown) revealed a moderate enhancement aftcr

administration of the contrast agent.

Findings o f the initial questionnaire. There was

no significant difference between B27-positive and B27-

negative subjects regarding back pain, joint pain, IBP,joint pain >30 days’ duration, and the combination of

back pain and joint pain (Table 4). Th e combination of

Table 4. Results of the questionnaire

N o . (%) N o . (5%)

B27+ B27-(n = 140) (n = 133) P

Back pain 104 (74.3) 92 (69.2) 0.40

Inflammatory back pain at night 32 (22.9) 19 (14.3) 0.I7Joint pain > ti days 75 (53.6) 68 (51.1) 0.75Eye inflammation 22 (15.7) 16 (12.0) 0.38Psoriasis 10 (7.1) 6 (4.5) 0.37Diarrhea >1 time/month 12 (8.6) 5 (3.8) 0.10

Inflammatory back pain 20154 (37.0) 6/40 (15.0) 0.01

Back pain and joint stiffness 8 (5.7) l(O.8) 0.026

Urethritis >1 time/year 7 (5.0) 1 (0.8) 0.036

and any 1 of the latter 4symptoms above




back pain and joint stiffness was more prevalent in the

B27-positive group. Also, the combination of back pain

and eye inflammation, back pain and urethritis >1

time/year, and back pain and any inflammation, but not

between back pain and diarrhea >1 time/year nor backpain and psoriasis, was more frequent in HLA-B27

positive subjects.

Th e overall frequency of back pain and joint pain

>8 days/year was high, but not significantly different

between the 2 groups. In B27-positive subjects, IBP at

night was more ofte n repo rted ; this was not significantly

different from that in B27-negative subjects (Table 4).

Eye inflammation, urethritis >1 time/year, diar-

rhea > I t ime/month, and psoriasis were re ported m ore

frequently in B27-positive donors. At least 1 of these

feat ures was found significantly more often in B27-

positive person s with IB P (20 of 54 [37%] versus 6 of 40HLA-B27 negative persons [15%]; P = 0.01).

Results of the personal medical history. T h e

characteristics of the 90 persons with possible SPA who

were examined (46 B27 positive, 44 B27 negative) did

not differ significantly from those of the total group.

These persons had a mean -+SD age of 39.5 ? 8.7 years,

70% were men, and all were white and of German

origin. Back pain was reported by 93.0% in the B27-

positive group and 92.7% in the B27-negative group,

69.8% and 63.4% had joint pain, 37.2% and 34.1% had

joint pain for >30 daysiyear, 23.3% and 26.8% rep orte d

joint swelling, and 14% and 12.2% had joint stiffness,respectively. Back pain at night waq significantly more

frequent in 18 of 46 B27-positive subjects (39.1%),

compared with 9 of 44 B27-negative subjects (20.5%).

Back pain in combination with eye inflammation was

more frequent, but not significantly so, in B27-positive

subjects than in B27-negative subjects (15.2% versus

6.8%), as was diarrhea >1 time/month (17.4% versus

9.1%; P = 0.25), urethritis >1 time/year (6.5% versus

2.3%; P = 0.34), and psoriasis (10.9% versus 6.8%;

P = 0.52).

In the 90 blood dono rs who were examined, ther e

was only a small difference between B27-positive and-negative persons w ho lost >3 days of work d ue to illness

over the preceding year (22.7% versus 18.2%; P not

significant). The 10 B27-positive donors had slightly

more time lost from work (me an ? SD 3 .1 2 2.5 weeks)

than the 8 B27-negative persons (2.2 + 2.4 weeks) ( P not

s ignif icant) . Eight of th e 20 b lood donors d iagnosed as

having SPA (40%) lost >3 days of work in the

previous year.

Findings of the physical examination. The w eight

and height of the 90 patients who underwent physical

examination did not differ between those who were B27

positive (79.4 kg, 176 cm) and those who were B27

negative (80.4 kg, 177 cm ). Anterio r spinal flexion,

measured according to Schober, an d thoracic expansion

were not significantly different between the 2 groups

(dat a not shown). Among th e 20 patients with a diagno-

sis of SPA, there were more persons who had thoracic

expansion of 5 3 cm (20% versus 14%) and more with

spinal flexion of 5 3 cm (40% versus 27%), but the

difference was not statistically significant.

DISCUSSION

The findings of this blood do nor study suggest a

high prevalence of spondylar throp athy. Nin eteen of 140

B27-positive blood donors (13.6%) assessed in this

study, but only 1 of 133 B27-negative donors (0.8%),fulfilled SPA criteria. Given a B27 frequency of 9.3%

among th e population of Berlin. a prevalence of SPA of

1.9% was calculated in this white population. These

numbe rs suggest that -45,000 adults between the ages

of 18 and 65 living in Berlin have SPA, two-thirds of

which are B27 positive.

The results of this first blood donor study per-

formed in whites to screen for SPA suggest that the

spondylarthropathies are the most frequent inflamma-

tory rheumatic diseases and tha t HLA-B27 is an impor-

tant risk factor. In comparison, rheumatoid arthritis, for

which the M H C class I1 molecule HLA-DR4 has beenidentified as a risk factor, has a lower prevalence of

-0.8-1.096 (35).

Th e SPA prevalence found in this study is in line

with recent work from the UK, where 5% of the patients

presenting to general practitioners with back pain were

diagnosed as having SP A (36). This percentage is fairly

high, given that back pain is a freq uen t complaint among

patients in primary care settings.

The overall SPA frequency reported herein is

slightly less than the 2.5% reported by Boyer and

colleagues in a study of an Eskimo p opulation >20 years

of age (15). In that study, USpA (1.3%) and Reiter’ssyndromeiKeA (1.0%) were more frequent than AS

(0.4%). The distribution of the S pA subsets in ou r study

is clearly different, with AS (0.86%) and U Sp A (0.67% )

being most frequ ent. T he higher prevalence of SPA

among E skimos is not surprising, since these po pulations

carry the B27 molecule by far more frequently than

Western Europeans, with prevalence rates of 20% and

more (17,18). However, th e reason for the differences in

disease expression between subsets is not clear. One of

the intriguing questions is which and how environm ental



64 BRAUN ET AL

factors, such as urogenital or enterobacterial infections,

influence disease expression (2). Or, from an othe r point

of view, which mechanisms protect Eskimos from devel-

oping AS.

Previous epidemiologic data on the spondylar-thropathies are mainly available for AS. The data from

our study point t o an AS prevalence of abou t 0.9%. This

is slightly less compared with the biggest population-

based study performed in Norway (21,329 persons stud-

ied), which was published in 1985 by Gra n and cowo rkers,

who found an AS prevalence of 1.1-1.4% (24). Again,

the difference between the 2 studies can be explained by

the different B27 frequency in the 2 populations (9.3%

versus 15.9% [37], respectively). However, o ur d ata are

i n contrast to a smaller population-based study from

Holland (2,957 persons studied), in which an AS preva-

lence of only 0.1-0.2% was found (23). Possible reasonsfor such a low prevalencc in that study have been

recently discussed (25). It should be stressed that in our

study, the responder rate of 85.3% was fairly high.

While in population-based studies, B27 is only

determined in cases, the AS prevalence was assessed in

a population of B27-positive subjects in this study. In a

blood d ono r study published 2 de cades ago (21), with a

design similar to ours, a very high AS prevalence of 18%

among B27-positive donors was found: 14 of 78 B27-

positive blood donors were reported to have pathologic

changes in the SI joints tha t were suggestive of S PA or

AS. In that study, pelvic radiographs, which have therep orte d problem of inter- and intra-observer variability

( 3 8 ) ,were used as the main diagnostic tool. In our study,

9 donors (6.4% of the B27-positive subjects) with defi-

nite SI changes were identified. Since we did not per-

form MRI in all persons, b u t rather, relied on clinical

history and specific symptoms to possibly identify a

wider spectrum of SPA patients, we cannot exclude the

possibility tha t persons with definite rad iograph changes

without clinical symptoms were m issed. Ou r approach of

screening for the presence of symptoms of interest might

give more useful clinical information. Nevertheless, if

th e 3 B27-positive donors with MRI-proven sacroiliitiswerc included, the prevalence of SPA in B27-positive

subjects would be 15.7%.

by M R I that was found in the present study was quite

high (5.8%). This can only be explained by the artificial

“enrichment” of B27-positivc subjects. Since our first

study published in Arthritis Kr RheLtmatism (28) and ou r

subsequcnt experiences with M R I for the detection of SI

changes, we have had no evidence of false-positive

findings.

The overall prevalence of sacro

Although no data on the prevalence of USpA in

whites have been published to date, we were not sur-

prised that th e second most frequent SPA diagnosis in

our study was USpA, since in a recen t consecutive study

of SPA patients who were referred to our outpatientclinic, US pA was the most frequent diagnosis (Braun et

al: unpublished observations). However, this might be

du e to referral bias, since we are screening especially for

patients in early stages of disease. The subset USpA

comprises SPA cases that d o not fit established diagnos-

tic categories of AS, ReA, PsA, and AIBD. USpA

patients were reported to be B27 positive in 70% of the

cases (1). Later in t he course of disease, a clearer clinical

picture of SPA might develop. The patients might show

overlapping and transient features of SPA or undiffer-

entiated disease might continue for years (39). This was

the case in 5 patients in ou r study who had a relativelylong history of symptoms >20 years, but no chronic SI

changes >grade 11. This result is consistent with data

from earlier studies in which SPA patients did not

develop progressive sacroiliitis or developed AS later,

after a mean of 9 years (39).

The overall prevalence of psoriasis in the popu-

lation is quite high: values of -2% are frequently cited

(40). According to th e answers to the que stionnaire, the

prevalence of psoriasis in Berlin would be almost 3 times

higher. We d o not believe tha t this is really th e case. It

just reflects the common clinical experience that the

majority of the patients, as well as many practitioners,used to call almost every squam atous skin change, such

as seborrheic d ermatitis, psoriasis. T hus, questionn aires

cannot be used to assess the prevalence of psoriasis in

the population. The patients diagnosed as having PsA

had psoriasis according to the diagnosis made by a

dermatologist. The reported percentage of psoriasis

patients with arthritis varies from 10% to 2096, suggest-

ing a PsA prevalen ce of 0.2-0.4% (41). In our study, 3

PsA patients were identified; this number would corre-

spond to an overall prevalence of 0.29% in the popula-

tion of Berlin. However, these nu mbers a re far too small

to draw conclusions. A more general problem in theclassification of PsA is tha t th ere ar e forms of PsA tha t

do not belong to the spondylarthropathies (42). At least

the subset resembling rheumatoid arthritis, which has a

frequency of 10-20% of all PsA patients, and the rarer

mutilating arthritis (3-596) should bc classified sepa-

rately. All patients identified in our study had the SPA

The identification of a patient with Reiter’s dis-

ea se was probably by chan ce and does not necessarily

ref lec t the t rue prevalence . Th e overa l l f requency of

type (1).




thir, subform of ReA was reported to be low (43),

while more benign forms of Re A seem to occu r more

frequently (44) .

The results of our initial questionnaire suggest

that IBP is not a valid instrument for screening for

spondylarthropathies in population-based studies, since

no difference between the B27-positive and B27-

negative groups was found. However, the persons we

examined were not “patients” who were actually com-

plaining about symptoms, and the value of a clinical

diagnosis of IBP might be different in a clinical setting,

where, according to our own experience and that of

others (12), IBP ir , useful for the initial evaluation of

SPA patients. H owever, IBP is not very specific (45), and

there are many early cases in which dynamic MRI, a

technique recently developed by our group, is useful for

demonstrating sacroiliitis on a more objective basis (28).

In the present study, 58 blood donors with IBP were

examined by M RI t o screen for signs of past or present

inflammation in the SI joints. M R I signs of different

degrees of sacroiliitis were found in 16 cases, 15 being

B27 positive. Thus, th e likelihood of having sacroiliitis in

a B27-positive individual with IBP can be calculated at

-SO%, while it is unlikely in B27-negative persons. We

conclude that a clinical diagnosis of IBP can be signifi-

cantly strengthened by positive determination of HLA-

B27 in SPA patients. Taking advantage of dynamic MR I

to a scerta in th e early diagnosis of sacroiliitis is clinically

helpful, especially in B27-positive patients. The rate at

which B27-positive patients with sacro

gressive sacroiliitis or other features typical of SPA has

not thus far been determined.

Although no extensive studies of the prevalence

of B27 in Germany have been performed, 2 arguments

can be put forward in w pp ort of the frequency of 9.3%

in the population of Berlin as reported in this study.

First, there was only slight variation between the 5 Berlin

blood banks involved in the study. Second, in older

studies from East Germany (46) and the south of

Germany (47), similar frequencies (12% and 8.9%,

respectively) in control groups have been reported.

Although B27 was the first ind epen den t variable

to discriminate between pa tients and controls, it was not

included in the ESSG classification criter ia ( l ) , because

classification criteria for epidemiolo gic studies should be

easy to assess. The B27-positive donors with MRI-

proven sacroiliitis identified in this study are likely to

have or to develop SpA. Thus, they could be included in

the USpA group. This would increase the percentage ofpatients with SPA am ong the B27-positive subjects and

make U SpA the most frequ ent SpA subset diagnosed in

this study.

Several methodologic aspects of this study need

to be discussed. In blood donor studies, an obvious

concern is referral bias, because blood donors are notfully representative of the population since not every-

body participates in blood donation. SP A patients with

severe disease would not be detected in our study

because they might not have been willing to give blood

or might not have been accepted as blood dono rs. Also,

children (48) and elderly people (49,50), in whom a

lower prevalence of SpA is assumed, were also no t

included in our study. However, spondylarthropathies

occur most frequently in the age group assessed. There

is also evidence tha t spondylitis can occur in the absence

of sacroiliitis in a small percen tage of cases ( 5 1,52); such

patients might have been missed. While the SPA preva-lence in the B27-positive group probably reflects the

situation in the general population, this is less clear in

the B27-negative group. E xtrapolating the only SpA case

of this group obviously leads to a huge variance. N one-

theless, the distribution of two-thirds in B27-positive

subjects and one-third in B27-negative subjects makes

sense when clinical experience is taken into account (1).

Furthermore, even if the number of the B27-negative

subjects were only half of the calculated figure, the

overall prevalence would still be 1.6%) because the

B27-positive group contributes the major part.

A main concern about population-based studiesis ascertainment bias. The criteria used in the present

study have been evaluated in European and North

American multicenter trials (5334). It is noteworthy that

the E SS G criteria have also been found useful in clinical

practice (55) . The 2 main ESSG criteria are inflamma-

tory spinal pain and synovitis ( I ) , the former defined

according to Calin ( 3 ) and the latter according to the

predominant localization in the lower limbs or the

asymmetric pattern of joint involvement. The duration

of arthritis, which is an important issue for the diagnosis

of rheumatoid arthritis, is not an essential part o f the

ESSG criteria. In this study, we took the presence orhistory of at least 6 weeks of joint pain and swelling with

no preceding t raum a as evidence of arthritis, while IBP

had to be present for at least 3 months (3).

What does an almost 2% prevalence of SpA

indicate? T he blood do nor s identified as SPA patients in

this study did not have very severe disease at presenta-

tion. but som e might later develop exacerbation of thcir

symptoms as a result of more severe inflammation or

increased bone formation and ankylosis. Accordingly,

Gran and colleagues compared AS patients identified in



BRAUN ET AL

a population-based survey (24) with patients in the

hospital. T he survey cases had somewhat l e s severe

disease (56). This makes sense and is consistent with ou r

results, as well as with t he physical examination findings,

in which no clear between-group differences concerning

chest expansion and spinal mobility could be detected.

Thus, this study contributes data on the prevalence of

SPA, especially in the B27-positive population, but as

discussed above, mo re severe cases were m ost probably

missed.

The disease course of SPA is variable, but is

possibly less severe tha n that of rheumatoid arthritis, as

judged from limited data (57) and clinical experience.

Prognostic markers have been proposed ( 58 ) ,but epide-

miologic data and long-term followup studies of such

patients are rare. N o increased mortality of AS patients,except for those who underwent radiotherapy, was recently

reported in the population of Rochester, MN (59).

The results of the questionnaire suggest a some-

what higher frequency of symptomatic urogenital and

ente ral inflammation in B27-positive subjects. This topic

has not been systematically assessed to date. Toivanen

and colleagues reported an increased frequency of anti-

bacterial antibodies in B27-positive blood donors (60).

Although it is tempting to speculate that B27-positive

persons have slightly deficient immune responses to

certain pathogc ns, which allow for longer persistence of

microbes and initiation and chronicity of arthritis andspondylitis, no evidence supporting this concept has so

far been provided. Our study suggests that larger

population-based studies might answer this question.

In conclusion, SPA and, especially, USpA are

frequent diseases that possibly run a chronic course.

Milder forms of SPA are often not diagnosed, partly

because of the rather nonspecific symptom of IBP.

HLA-B27 is an impo rtant risk factor for SPA. A positive

B27 result in patients with IBP early in the disease

course is suggestive of sacroiliitis and might indicate

possible progression to the development of SPA.

ACKNOWLEDGMENTS

We acknowledge the essential cooperation of Dr. E.

Keller (Deutsches Ro tes Krcuz, Klinikum Berlin Buch), Dr. C.

SchGnemann (Krankenhaus Berlin-Friedrichshain), Dr. J .

Siaplaouras (Klinikum Rudolf Virchow der Hu mbol dt Univer-

sitat), and Dr. M. Bunte (Departmcnl of Hematology, Klini-

kum Benjamin Franklin der Freien Universitat Berlin), who

contributed the addresses and HLA typing results of the blood

donors, with the donors’ consent.

REFERENCES

1. Dougados M, van der Linden S, Juhlin R, Huitfeldt B, Amor B,Calin A, ct al. The European Spondylarthropathy Study Grouppreliminary criteria for the classification of spondylarthropathy.

Arthritis Rheum 1991;34:1218-27.2. Zeidler H, Mau W , Khan MA. Undifferentiated spondyloar-

thropathies. Rheum Dis Clin North Am 1992;18:187-202.3. Calin A, Porta .I, Fries J, Schurman DJ. Clinical history as a

screening tcst for ankylosing spondylitis. JAMA I977;237:26 13-4.4. Bellamy N, Park W, Rooney P. What do we know about the

sacroiliac joint. Semin Arthritis Rheum 1983;12:282-3 13.

5. Braun J, Sieper J . The sacroiliac joint in the spondyloarthropa-thies. Curr Opin Rheumatol 1996;8:275-87.

6 . Brewerton DA, Caffrey M, Nicholls A, Walters D, Oates JK,James DC. Reiter’s disease and HL-A 27. Lancet 3973;2:996-8.

7. Brcwcrton D, Hart F, Nicholls A, Caffrey M, James DC, SturrockRD. Ankylosing spondylitis and HLA27. Lancet 1973;1:904-7.

8. Rubin LA, Amos CI, Wade JA, Martin JR, Bale SJ , Little AH, etal. Investigating the genetic basis for ankylosing spondylitis: link-age studies with the major histocompatibility complex region.

Arthritis Rheum 1994;37:1212-20.9. Sieper J, Braun J. Pathogenesis of spondylarthropathies: persistent

bacterial antigen, autoimmunity, or both? Arthritis Rheum 1995;

38:1547-54.10. Linssen A, Feltkamp TE. B27 positive diseases versus B27 nega-

tive diseases. Ann Rheum Dis 1988;47:431-9.11 . Khan MA, Khan MK. Diagnostic value of HLA-B27 testing in

ankylosing spondylitis and Reiter’s syndrome. Ann Intern Med1982;96:70-6.

12. Dougados M. Diagnostic fcatures of ankylosing spondylitis. Br J

Rhcumatol 1W5;34:3O1-5.13. &an JT, Husby G. HLA B2 7 and spondyloarthropathy: value for

early diagnosis? J Mcd Genet 1995;32:497-501.14 . Khan MA. HLA-B27 and its subtypes in world populations. Curr

Opin Rheumatol 1995;7:263-9.15 . Boyer GS, Templin DW, Cornoni-Huntley JC, Everett I)F, Law-

rence RC, Heyse SF, et al. Prevalence of spondyloarthropathies in

Alaskan Eskimos. J Rheumatol 1994;21:2292-7.16, Alcxccva L, Krylov M, Vturin V, Mylov N, Erdesz S, Benevolen-

skaya L. Prevalence of spondyloarthropathies and HLA-B27 in thenative population of Chukotka, Russia. J Rheumatol 1994:21:2298-300.

17 . Hansen JA. Lanicr AP, Nisperos B, Mickelson E, Dahlberg S. The

HLA system in Inupiat and Central Yupik Alaskan Eskimos. HumTmmunol 1986;16:315-28.

18. Khan MA, Kellner H. Immunogenetics of spondyloarthropathies.Rheum Dis Clin North Am 1992;18:837-64.

1Y . Gofton JP. Epidemiology, tissue type antigens and Bechterew’ssyndrome (ankylosing spondylitis) in various ethnic populations.Ann Rheum Dis 1966;25:525-7.

20 . Gofton JP, Bennett PH, Smythe HA, Decker JL. Sacroiliitis andankylosing spondylitis in North American Indians. Ann Rheum

Dis 19723 :474-81.21. Calin A, Fries J. Striking prevalence of ankylosing spondylitis in

‘healthy’W27 positive males and females. N Engl J Mcd 1075;293:835-9.

22. Cohcn LM, Mittal KK, Schmid FR, Rogers LF, Cohen KL.Increased risk for spondylitis stigmata in apparently healthy HL-AW27 men. Ann Intern Med 1976;84:1-7.

23. Van der Linden S, Valkenburg HA, de Jongh BM, Cats A. Therisk of devcloping ankylosing spondylitis in HLA-B27 positiveindividuals: a comparison of relatives of spondylitis patients withthe general population. Arthritis Rheum 1984;27:241-9.

24. Gran JT, Husby G, Hordvik M. Prevalence of ankylosing spondy-litis in males and females in a young middle-aged population ofTromso, northern Norway. Ann Rheum Dis 1985;44:359-67.




25. Gran JT, Husby G. The epidcmiology of ankylosing spondylitis.Semin A rthritis R heum 1993;22:319-34.

26. Khan M A, van de r Linden SF . Th e risk of ankylosing spondylitis inHL A B27 positive individual reappraisal. J Rhe uma tol 1984;11:727-8.

27. LeClercq SA, Russell AS. The risk of sacroiliitis in B27 positivepersons: a reappraisal. J Rheumato l 1984;11327-9.

28. Braun J, Bollow M, Eggcns U, Kiinig 11, Distler A, Siepcr J. Useof dynamic magnetic resonance imaging with fast imaging in thedetection of early and advanced sacroiliitis in spondylarthropathypatients . Arthritis Rheum 199437:1039-45.

29. Bollow M , Braun J, Ham m B, Eggens U , Schilling A, K onig H, e tal. Early sacroiliitis in patients w ith spondy loarthropathy: evalua-tion with dynamic gadolinium-enhanced MR imaging. Radiology1995;194:529-36.

30 . Braun J , Bollow M, Seyrekbasan F, Hibcr lc HJ , Eggens U , MertzA, et al. Com puted tom ography guided co rticosteroid injection ofsacroiliac joint in pa tients with spondyloarthropathy with sacroili-itis: clinical outcom e and followup by dynamic ma gnetic resonanceimaging. J Rheumato l 1996;23:659-64.

31. Schober P. Lendenwirbelsaule und Kreuzschmerzen. Munch Med

Wschr 1937;84:336-428.32. Moll JM H, Wrigh t V. An objective clinical study of chest expan-

sion. An n Rh eum Dis 1972;31:1-8.33. Van der Linden S, Valken Cats A. Evaluation of diagnostic

criteria for ankylosing spo proposal for modification of theNew York criteria. Arthritis Rh eum 1984;27:361-8.

34. Willkens RF , Arnett FC , Bittcr T, Calin A , Fisher L, Ford DK, etal. Reiter’s syndrome: evaluation of preliminary criteria for dcfi-nite disease. Arthritis Rheum 1981;24:844-9.

35. Alarcdn GS. Epidemiology of rheumatoid arthritis. Rheum DisClin North Am 1995;21:589-604.

36. Undewood MR, Dawes P. Inf lammatory back pain in pr imarycare. Br J Rheum atol 1995;34:1074-7.

37. Gran JT, Mcllbye AS, Husby G. Thc prevalence of HLA-B27 innorthern Noway. Scand J Rheumatol 1984;13:173-6.

38. Hollingsworth PN, Cheah PS, Dawkins RL. Owen ET, Calin A,

Wood PHN. Observer variation in grading sacroiliac radiographsin HL A-B27 positive individuals. J Kheumatol 1983;10:247-54.39. Mau W, Zcidlcr H , Mau R. Majewski A, Freyschmidt J, Stangel

W, e t al. Clinical features and prognosis of patients with possibleankylosing spondylitis: results of a 10-year followup. J Rhcumatol

40. Swanbcck G, Inerot A, Martinsson T, Wahlstrom J. A populationgenetic study of psoriasis. Br J Dermatol 1994;131:32-9.

41. O’Neill T, Silman AJ. Psoriatic arthritis: historical background andepidemiology. Baillieres Clin R heum atol 1994;8:245-61.

42. Salvarani C, Lo Scocco G, Macchioni P, Cremoncsi T, Rossi F,Mantovani W. Prevalence of psoriatic arthritis in Italian psoriaticpatients. J Rheumatol 199S;22:1499-15O3.

43 . IsomakiHA .An epidemiologically based follow-up study of recentarthritis: incidence, outcome and classification. Clin Rheumatol1987;6 Suppl 253-9.

44 . Glennas A. Kvien TK, Melby K, Overboo A . An d r u p 0,

Karstensen B, et al. Reactive arthritis: a favorable 2 ycar course

1988;15:1109-14.

and outcome, independent o f t r igger ing agent and HIAB27.J Rheumatol I 94;21:2274-80.

45 . Blackburn WD Jr, Alarccin GS, Ball GV. Evaluation of palicntswith back pain of suspected inflammatory nature. Am J Med

46 . Wcgener VS, Mach J . Zu r Assoziation von ankylosierender SpoJl-dylitis und HLA-System. Z Arztl Forthild (Jena) 1980:74:895-8.

47. Wenk HF , Ruprecht KW, Schunter F. FlLA antigens in anterioruveitis. Albrecht Von Graefes Arch Klin Exp Ophthalmol 1980;

215:137-43.

48. Burgos-Vargas R, V izyuez-Mellado J. The early clinical recogni-tion of jwcn ile-onse t ankylosing spondylitis and its differentiationfrom juvenile rheumato id arthritis. Arthritis Rhcum l995;38:835-

44.49. Dubost JJ, Sauvezie B. Latc onset peripheral spondyloarthro-

pathy. J Rheumatol 1989;16:1214-17.50 . Olivieri I, Padula A, Pierro A , Favaro L, Oranges GS, Ferr i S. La tc

onset undiflferentiatcd seronegative spondyloarthropath y. J R heu-matol 1995;22:899-903.

51 . Gran JT, Husby G, Hordvik M. Spinal ankylosing spondylitis: a

variant form of ankylosing spondylitis or a distinct disease entity’?Ann Rheum Dis 1985;44:368-71.

52. Khan MA, van der Lindcn SM, Kushner I, Valkenburg HA, CatsA. Spondylitic disease without radiologic evidence of sacroiliitis in

relatives of HLA-B27 positive ankylosing spondylitis patients.Arthritis Rheum 1985;28:40-3.

53. Collantes Estcvez E, Cisnal del Mazo A, Munoz Ciomariz E.

Assessment of 2 systems of spondyloarthropathy diagnostic andclassification critcria (Amor an d ES SG) by a Sp anish multicentcrstudy: European Spondyloarthropathy Study Group. J Rhcurnatol1995;22:246-5 I .

54. Boyer GS, Templin DW, G oring WP. Evaluation of the EuropcanSpondylarthrop athy Study G rou p preliminary classification crite-

ria in Alaskan Eskimo populations. Arthritis Rheum 1 9 9 3 3 6 :

534-8.55. Amo r B, Dougados M, Listrat V, Menkes CJ, Roux H, Bcnhamou

C, et al. Are classification criteria for spondylarthropathy uscful asdiagnostic criteria? Rev Rhum Engl Ed 1995;62:10-5.

56 . Carette S, Graham D, Little H, Rubenstein J, Roseri P. Thenatural disease course of ankylosing spondylitis. Arthritis Rheum1983;26:186-90.

57. Gran JT, Husby G . Ankylosing spondylitis: a com parativc study ofpatients in an epidemiological survey, and those admitted to adepartment of rheumatology. J Rheumato l 1984;1 1 :788-93.

58. Amor B, Santo s RS. Nahal R, Listrat V, Dougados M. Predictivefactors for thc long-tcrin outcome of spondyloarthropathies.J Rheumatol 1994;21:1883-7.

59 . Carbone LD, Coopcr C, Michct CJ, Atkinson EJ, O’Fallon WM,Melton LJ 111. Ankylosing spondylitis in Rochcster , M innesota,J 935-1989: is the epidemiology changing? Arth ritis Rh eum 1992;35:1476-82.

00. Toivanen P, Koskimies S, Granfors K, Eerola E. Bacterial anti-bodies in HLA-B27+ healthy individuals. Arthritis Rheum 1993;36:1633-5.

1988;85:766-70.

Prevalence of Spondylarthropathies

Documents

Transcript of Prevalence of Spondylarthropathies