Original Paper

Neuroepidemiology 2003;22:82–86DOI: 10.1159/000067101

The Frequency of Multiple Sclerosis inJewish and Arab Populations in GreaterJerusalem

Arnon Karnib Esther Kahanaa,c Nelly Zilberc,d Oded Abramskyb

Milton Altere Dimitrios Karussisb

aDepartment of Neurology, Barzilai Medical Center, Ashkelon, bDepartment of Neurology,Hadassah Hebrew University Hospital, Hadassah Medical School, cNeuroepidemiology Unit,Hadassah Hebrew University Hospital, dC.N.R.S., Centre de Recherche Français de Jérusalem, Jerusalem, Israel;eDepartment of Neurology, MCP/Hahnemann University, Philadelphia, Pa., USA

Dr. Arnon KarniCenter for Neurologic DiseasesBrigham and Women’s Hospital, Harvard Medical School77 Avenue Louis Pasteur, Boston, MA 02115 (USA)Tel. +1 617 525 5184, Fax +1 617 525 5252, E-Mail akarni@rics.bwh.harvard.edu

ABCFax + 41 61 306 12 34E-Mail karger@karger.chwww.karger.com

© 2003 S. Karger AG, Basel0251–5350/03/0221–0082$19.50/0

Accessible online at:www.karger.com/ned

Key WordsMultiple sclerosis W Immigrants W Environmental factors W

Arabs W Jews W Israel

AbstractA comparison of the incidence rate (IR) and the preva-lence rate (PR) of multiple sclerosis (MS) in subgroups ofthe same ethnic origin, but born and living in differentgeographical areas, may delineate the relationship be-tween environmental and genetic risk factors for MS.Previous epidemiological studies of MS in Israel did notinclude the Arab population and used diagnostic criteriathat did not include MRI findings. Therefore, we studiedthe age-adjusted IR and PR of MS in a more recent sam-ple in different population groups, including Arabs, ofGreater Jerusalem. We found that the PR of MS in IsraeliJews is higher than previously described. Furthermore,the PR was significantly lower among immigrant Jewsfrom Asia/Africa (A/A) than among native-born Jews ofAsian/African origin (I-A/A). Since these groups havesimilar genetic susceptibilities to MS, the higher PR inthe latter is probably due to environmental factors. Our

study does not support the effect of latitude on the risk ofdeveloping MS since no difference in the PR was foundbetween immigrant Jews from Europe/America (E/A)and native-born Jews of European/American origin(I-E/A). Among Arabs, the PR was similar to that amongA/A. Therefore, we hypothesized similarity in environ-mental etiologic factors for MS between the countries oforigin of A/A immigrants and of Arabs communities inGreater Jerusalem. The IR of I-E/A was higher than thatof I-A/A and Arabs, although this difference did not reachstatistical significance.

Copyright © 2003 S. Karger AG, Basel

Introduction

Multiple sclerosis (MS) is a chronic white-matter dis-ease of the central nervous system, that affects mainlyyoung adults, the etiology of which is still unknown [1].The uneven geographical distribution of MS prevalencehas long been thought to offer an opportunity to study thepossible role of genetic and environmental factors pre-sumed to be of etiological importance [2]. Comparisons of

MS in Greater Jerusalem Neuroepidemiology 2003;22:82–86 83

disease frequency in immigrants from regions differing inMS frequency, their offspring born and reared in a newenvironment and the native-born host populations are ofparticular interest in this regard. Immigrants and theiroffspring may change their environmental milieu, whileremaining genetically alike. Thus, if MS frequency re-mains the same in the immigrant population and their off-spring (born and reared in a different environment),genetic factors in the etiology are favored. On the otherhand, if the offspring show a different frequency of MSthan their immigrant parents and come to resemble thenative-born host population, then environmental factorsare implicated in the etiology.

A number of studies in different regions have demon-strated that immigrants who moved from regions withhigh rates of MS to areas with lower MS rates, e.g. fromthe United Kingdom to South Africa before the age of 15[3] or from Washington State to California [4], decreasedtheir rate of MS. On the other hand, populations thatmigrated from low MS regions to regions where MS iscommon (e.g. from the West Indies and from Africa to theUK) [3, 5] retained their low rate of MS. However, theoffspring of the West Indian and African immigrants tothe UK, born in the new environment, showed a higherrate of MS, similar to that of the host population [6].

Opportunities to study the effects of genetic and envi-ronmental factors in MS etiology would be enhanced ifthe host population in a given region had identifiable sub-populations with diverse MS frequency as well as immi-grant groups who arrived from regions with different MSfrequencies. Israel is such a country. Since 1948, immi-grants arrived in Israel in large numbers from differentregions, some with high rates of MS (e.g. Europe andNorth America) and some with low MS rates (e.g. NorthAfrica and Asia). The native-born Israeli population in-cludes Jews who have lived in the region for generations,the offspring of recent Jewish immigrants as well as Arabinhabitants of Israel and their offspring. This diversity,coupled with the fact that medical care is of high qualityand readily accessible to all of Israel’s inhabitants, made itan excellent venue for epidemiological studies designed toinvestigate environmental and genetic factors in the etiol-ogy of MS.

Previous epidemiological studies in Israel, the first ofwhich was reported in 1962 [7], showed that in immi-grants from Europe, the MS prevalence rate ranged from30 to 51/105 population, while in immigrants from NorthAfrica and Asia (A/A), the prevalence ranged from 3 to7/105. Although these estimates of frequency were inter-preted to indicate that European Jewish immigrants had a

higher prevalence of MS than the African-Asian immi-grants, they were crude estimates that did not take intoaccount the differences in age structure of the groups. TheEuropean immigrants, for example, were older on averagethan the A/A. Because MS usually first appears after ado-lescence, the older European immigrant populationwould be expected to have more patients among whomMS had already been recognized. Moreover, the earlystudies of MS in Israel included clinically probable andpossible MS cases in the estimates and did not have mod-ern diagnostic tools such as MRI available.

A more recent study in Israel [8] calculated age-adjusted MS prevalence rates and used the diagnostic cri-teria of McDonald and Halliday [9], Bauer [10] and Roseet al. [11]. A review of the cases included in that studyrevealed that age adjustment and current diagnostic crite-ria, e.g. those of Poser et al. [12], did not change the rela-tive MS prevalence in the Jewish groups. The latter studyestimated MS prevalence in the greater Jerusalem area.MS prevalence in native-born Jews with European-Amer-ican ancestry (I-E/A) in this area was 38.5/105, and innative-born Jews with African-Asian forebears (I-A/A),the prevalence rate was almost the same (31.1/105). Incontrast, in A/A, the prevalence of MS was 16.3/105, i.e.only half as high. These observations pointed to factors inthe Israeli environment that increased the risk of MS.

In the present study, we calculated MS prevalence andincidence rates in the greater Jerusalem area not only inJewish but also in Arab inhabitants. In Israel, Arabs andJews share the same geoclimatic environment but havedifferent lifestyles (e.g. food types, family size, occupa-tions). However, prior to their immigration to Israel, theJews from Arab countries of North Africa and Asia had alifestyle similar to that of the Israeli Arabs. Therefore, ifenvironmental factors, especially those to which individu-als are exposed early in life, influence MS frequency,Arabs and Jewish immigrants from Arab countries wouldbe expected to have similar rates of MS.

The available data on MS frequency in Arab countries,e.g. Jordan [13], Kuwait [14] and Saudi Arabia [15], indi-cate that this disease is uncommon. While the presentstudy, which includes the Arab population, offers anopportunity to reassess the possible influence of environ-mental factors associated with living in Arab communi-ties, it does not exclude the role of genetic factors either.Environmental factors, however, may be more amenableto modification and, thus, offer greater hope of modifyingthe risk of MS.

84 Neuroepidemiology 2003;22:82–86 Karni/Kahana/Zilber/Abramsky/Alter/Karussis

Patients and Methods

The Study AreaGreater Jerusalem is a designated, geographically defined area for

which census data are available, with information on the number ofinhabitants and their ethnic composition as well as their regions oforigin. Greater Jerusalem includes: the city of Jerusalem, 1 town, 2suburbs, 12 communal settlements (kibbutzim) and 48 villages, allwithin a 40-km distance from Jerusalem.

Patient AccrualThe Hadassah University Hospital is the only medical center in

the area of Greater Jerusalem that has a Department of Neurology.This hospital has modern MRI facilities, a laboratory for CSF analy-ses and electrophysiological equipment for evoked potential studies.The Department of Neurology is heavily oriented towards neuroim-munology, with expertise in MS. It is a tertiary neurological center towhich patients with neurological disease are regularly referred fromall over Israel. A National MS Register was established in 1960 byone of us (M.A.) and has been maintained with periodic updatessince then.

We reviewed the medical records at the Hadassah Hospital aswell as the records of all other hospitals (e.g. Shaare Tzedek) andmedical clinics in Greater Jerusalem for patients with a diagnosis ofMS or a neurological disorder that could be MS (e.g. optic neuritis,paraparesis, cerebellar ataxia, lateral sclerosis or Devic’s neuromyeli-tis optica). We also reviewed the computerized records of the Nation-al MS Register. Finally, the Israeli MS Society cooperated in the caseascertainment. All MS cases identified were classified as to certaintyof diagnosis using the criteria by Poser et al. [12].

Population at RiskThe population at risk on the Prevalence Day, i.e. the number of

residents in Greater Jerusalem on December 31, 1995, was providedby the 1996 Yearbook of the Israeli Central Bureau of Statistics,which is based on the national census [16]. The census data areupdated annually, taking into account births, deaths, immigrationand emigration. Data are recorded by region and include informa-tion on age, sex, religion, country of birth and, for those born inIsrael, country of birth of the father. Although there is some intermar-riage between different ethnic Jewish groups, it has been estimated at!10%, i.e. the father’s country of origin reflects the ethnic origin ofthe Israeli-born group with reasonable accuracy. Intermarriage be-tween Jewish and Arabs is even lower. In our study, we included allMoslem Arabs but not the Christian Arabs because the Israeli Cen-tral Bureau of Statistics provides census data by religious affiliationof the population, and many Christian inhabitants in Greater Jeru-salem are not Arabs. Moreover, Moslem Arabs constitute the largestgroup of Arabs in Israel as well as in Greater Jerusalem. The origin ofIsraeli Jews whose father was born in Israel was estimated using theproportion of residents with Europe/America or Asia/Africa origin inthe census data of 1960 and 1971 [17, 18].

Incidence and Prevalence RatesWe calculated the age-adjusted prevalence rate of MS (direct

adjustment) for each of the study groups, i.e. immigrant Jews fromEurope or America (E/A), A/A, I-E/A, I-A/A and Arabs. In order tocompare the results of the present study with previous studies of MSfrequency in Israel, we used the age structure of the Israeli populationin 1960 as a standard for the age-adjusted rates since this was the

Table 1. Age-adjusted prevalence rates in the study populationgroups

Populationgroup

Populationat risk

MScases

AdjustedPR B SE1

E/A 114,966 88 64.3B7.1A/A 54,945 21 22.1B5.9I-E/A 137,458 65 63.5B8.2I-A/A 202,872 79 52.1B7.9Arabs 167,300 19 19.2B6.2

PR = Prevalence rate, E/A = European/American Jews, A/A =African/Asian Jews, I-E/A = Israel-born European/American Jews,I-A/A = Israel-born African/Asian Jews.1 Age-adjusted prevalence rate per 105 inhabitants (direct adjust-ment using the 1960 Israeli population as a standard) on December31, 1995 and B standard error of the rate.

standard population used in previous studies of MS frequency inIsrael.

Data were collected during 1998. The Prevalence Day for thepresent study was December 31, 1995. The average annual incidencewas calculated based on the new cases of MS that had their onset inIsrael between July 1, 1994 and June 30, 1997.

StatisticsConfidence intervals (95%) for estimates were calculated assum-

ing a Poisson distribution.

Results

The total population in Greater Jerusalem that wasincluded in the present study was 677,541 (table 1). Onthe Prevalence Day, December 31, 1995, there were 272definite or probable MS patients who were living as resi-dents of Greater Jerusalem. Of these, 169 were femalesand 103 were males (ratio 1.64:1, F:M). A total of 253were Jews and 19 were Arabs. All of the Arab patientswere born in Israel. Table 1 indicates the distribution bypopulation group of the MS cases, the population at riskand the age-adjusted prevalence rates of MS in each of thepopulation groups. As is apparent in this table, the preva-lence of MS is very similar in I-E/A and in E/A (amongI-E/A 63.5/105 population and among E/A 64.3/105). Incontrast, the A/A who were born and lived in an Arabcountry prior to immigration had a much lower preva-lence of MS, which was virtually the same as the preva-lence of MS in Israeli-born Moslem Arabs (among A/A22.1/105 and among Arabs 19.2/105). The I-A/A who wereborn and grew up in an Israeli-Jewish environment had a

MS in Greater Jerusalem Neuroepidemiology 2003;22:82–86 85

prevalence rate of MS of 52.1/105, which is significantlyhigher than the prevalence of MS in their immigrant fore-bears.

The mean age-adjusted annual incidence rate amongIsraeli-born inhabitants of Greater Jerusalem were calcu-lated for the 3-year period between July 1, 1994 and June30, 1997. The rate was higher among the I-E/A (3.1/105

inhabitants, SD = 2.9) than among the I-A/A (1.3/105,SD = 1.2) and the Arabs (0.7/105, SD = 1.1). This trendwas similar to that observed with the prevalence rates; butperhaps because the numbers on which these incidencerates were based were small, the differences between thegroups did not reach statistical significance.

Discussion

The frequency rates of MS reported in the presentstudy in Israel are higher than those reported previously.The question of whether such an increase is real or an arti-fact has been discussed by Riise [19]. The advent of moresensitive diagnostic tools such as MRI in Israel has cer-tainly played a role in identifying cases as ‘definite MS’that would previously have been considered ‘possible MS’and, thus, excluded from some frequency calculations.Sensitive diagnostic tools may also shorten the intervalbetween first symptoms and the time of diagnosis, whichincreases the known cases of MS and therefore the preva-lence rates. The quantity and quality of health care facili-ties has increased and in particular the neurological train-ing of physicians has improved in Israel over the years, sothat patients with symptoms of MS now have a greaterchance of being diagnosed correctly as MS. Therefore, it islikely that the observed increase in MS frequency in Israelis more apparent than real, but in truth, the questionremains open.

Genetic characteristics that could influence the risk ofdeveloping MS remain the same among both immigrantsand offspring born in the adopted country. Some environ-mental factors are similar among the different studygroups, e.g. geoclimatic factors, while other environmen-tal factors may be different, e.g. exposure to pathogens,utilization of health care resources, food types, habitat,family size and occupation, which all tend to shift afterimmigration to resemble the patterns in the new abode.

Our study is the first in Israel to make a special effort toidentify Arab patients with MS. It is reasonable to assumethat genetic differences exist between Arabs and Jews inIsrael, but the genetic characteristics of Arabs with MS,such as the HLA markers are, as yet, not well defined [20].

The genetic issue, as at least a partial explanation for thedifference in frequency of MS between Arabs and Jews inIsrael, requires more information before resolution is pos-sible. However, the observation that A/A and I-A/A havesignificantly different prevalence rates of MS pointsstrongly toward an environmental etiological factor. Thisinference is further strengthened by the fact that amongA/A, who are mostly immigrants from Arab countries (e.g.Iraq, Morocco, Syria, Egypt and Yemen) or from coun-tries with a Moslem majority (e.g. Iran, Uzbekistan, Ka-zakhstan, Tajikistan, Azerbaijan, Turkmenistan and Kyr-gyzstan), the prevalence rate of MS is probably similar tothat among Moslem Arabs in Israel. The I-A/A, contraryto the expected predominant genetic risk factors, haverates approximating those of E/A and those of I-E/A.Thus, in a single generation living in an environment thatis more like Europe and America than like that of Arabcountries, the prevalence of MS in the I-A/A has approxi-mated the higher ‘European/American’ rate. These obser-vations appear to implicate environmental factors in theIsraeli lifestyle in increasing the risk of MS or, alternative-ly, environmental factors in the lifestyle of Arab countriesin protecting against developing MS. Furthermore, thereported crude rates from other Arab populations [13–15]are generally low and similar to the rate found in ourstudy for the Arab population. However, the rates of MSamong the residents of Palestinian origin in Jordan [13]and Kuwait [14] are higher than the native-born popula-tion in these countries. These rates are in agreement withour hypothesis of a protective factor in the Arab lifestyle.However, since the rates are not age-adjusted and the agestructures of the Jordanian, Kuwaiti, Saudi Arabian pop-ulations and the populations of the Arabs in our studymay be quite different, these comparisons may be prob-lematic.

One environmental factor that might explain the ob-served differences in MS frequency among the variousgroups studied is a difference in age at exposure to child-hood infections [21, 22]. We postulate early exposure tochildhood infections in Arabs in Israel, and in A/A priorto leaving their country of origin. The I-A/A, E/A and I-E/A are postulated to have had a later exposure. Such anexplanation is attractive because it is consistent with thehypothesis that MS may be due to altered ‘fine-tuning’ ofthe immune system by the age at which infection occurs[23]. According to this hypothesis, earlier exposure, suchas occurs in Arab countries, may have ‘protected’ theimmigrants from these countries from MS to some de-gree. Their offspring born and reared among Jews in Israelwere less protected, as Jews in Israel have a more ‘West-

86 Neuroepidemiology 2003;22:82–86 Karni/Kahana/Zilber/Abramsky/Alter/Karussis

ern’ lifestyle. Therefore, the MS rate is higher in theseIsraeli-born offspring than in their immigrant forebears.

This interpretation of the data in the present study issomewhat speculative, not only because very little isknown about the age pattern of childhood infections inJewish and Arab groups in Israel, but also because manyother factors differ in the various groups studied. A case-control study in the different study groups might help todisclose the relevant risk factors [24]. Meanwhile, our epi-demiological study of MS frequency in Jewish and Arabinhabitants of Greater Jerusalem already indicates thatenvironmental factors are significantly implicated in MSetiology. It will be of interest to follow the trend in fre-quency of MS among the groups studied because Israel,

already very Western in its lifestyle, continues to developand the ‘Western’ lifestyle also influences the Arab popu-lation.

Acknowledgments

We thank Dr. Padmanabhan Bharanidharan for helpful discus-sions and suggestions and the Israeli MS Study Group (Drs. R. Bloch,J. Chapman, S. Flechter, R. Gilad, Y. Herishanu, I. Koren, E. Kot, A.Miller, R. Milo, R. Mosberg-Galili, P. Nisipeanu, L. Rasin, R. Sha-hien, W. Simri, B. Weller, D. Yarnitski) for their help in updating theMS registry. This work was supported in part by the Hilda KatzBlaustein, the Zeev Aram and the Lena P. Harvey EndowmentFunds. The registry was supported by Teva Pharmaceutical Indus-tries Ltd., Israel.

References

1 Karni A, Weiner HL: Organ-specific inflam-matory disease: Multiple sclerosis; in Rich RR,Fleisher TA, Shearer WT, Kotzin BL, SchroedeHW Jr (eds): Clinical Immunology. Principlesand Practice. London, Mosby, 2001, pp 77.1–77.12.

2 Rosati G: Descriptive epidemiology of multi-ple sclerosis in Europe in the 1980s: A criticaloverview. Ann Neurol 1994;36(Suppl 2):S164–S174.

3 Dean G, Elian M: Motor neuron disease andmultiple sclerosis mortality in Australia, NewZealand and South Africa compared with En-gland and Wales. J Neurol Neurosurg Psychia-try 1993;56:633–637.

4 Visscher BR, Detels R, Coulson AH, Malm-gren RM, Dudley JP: Latitude, migration andprevalence of multiple sclerosis. Am J Epide-miol 1977;106:470–475.

5 Kurtzke JF: Multiple sclerosis among immi-grants. Br Med J 1976;1:1527–1528.

6 Elian M, Dean G: Multiple sclerosis among theUnited Kingdom-born children of immigrantsfrom West Indies. J Neurol Neurosurg Psychia-try 1987;50:327–332.

7 Alter M, Halpern L, Kurland LT, Bornstein B,Leibowitz U, Silberstein J: Multiple sclerosis inIsrael: Prevalence amongst immigrants and na-tive inhabitants. Arch Neurol 1962;7:15–25.

8 Kahana E, Zilber N, Abramson JH, Biton V,Leibowitz Y, Abramsky O: Multiple sclerosis:Genetic versus environmental aetiology: Epi-demiology in Israel updated. J Neurol 1994;241:341–346.

9 McDonald WI, Halliday AM: Diagnosis andclassification of multiple sclerosis. Br Med Bull1977;33:4–9.

10 Bauer HJ: IMAB-enquete concerning the diag-nostic criteria of multiple sclerosis; in BauerHJ, Poser S, Ritter G (eds): Progress in Multi-ple Sclerosis Research. Berlin, Springer, 1980,pp 555–563.

11 Rose AS, Ellison GW, Myers LW, TourtellotteWW: Criteria for the clinical diagnosis of mul-tiple sclerosis. Neurology 1976;26:20–22.

12 Poser CM, Paty DW, Scheinberg L, McDonaldWI, Davis FA, Ebers GC, Johnson KP, SibleyWA, Silberberg DH, Tourtellotte WW: Newdiagnostic criteria for multiple sclerosis:Guidelines for research protocols. Ann Neurol1983;13:227–231.

13 Najim Al-Din AS, Kurdi A, Mubaidin A, El-Khateeb M, Khalil RW, Wriekat AL: Epidemi-ology of multiple sclerosis in Arabs in Jordan:A comparative study between Jordanians andPalestinians. J Neurol Sci 1996;135:162–167.

14 Najim Al-Din AS, Khogali M, Poser CM, et al:Epidemiology of multiple sclerosis in Arabs inKuwait: A comparative study between Kuwai-tis and Palestinians. J Neurol Sci 1990;100:137–141.

15 Al Rajeh S, Bademosi O, Ismail H, Awada A,Dawodu A, al-Freihi H, Assuhaimi S, BorollosiM, al-Shammasi S: A community survey ofneurological disorders in Saudi Arabia: TheThugbah study. Neuroepidemiology 1993;12:164–178.

16 Israeli Central Bureau of Statistics (1996). Sta-tistical Abstract of Israel.

17 Israeli Central Bureau of Statistics (1972). Sta-tistical Abstract of Israel.

18 Israeli Central Bureau of Statistics (1961). Sta-tistical Abstract of Israel.

19 Riise T: Is the incidence of multiple sclerosisincreasing?; in Thompson AJ, Polman C, Hohl-feld R (eds): Multiple Sclerosis: Challenges andControversies. London, Dunitz, 1997, pp 1–12.

20 Kwon OJ, Karni A, Israel S, Brautbar C, AmarA, Meiner Z, Abramsky O, Karussis D: HLAclass II susceptibility to multiple sclerosisamong Ashkenazi and non-Ashkenazi Jews.Arch Neurol 1999;56:555–560.

21 Alter M, Zhang ZX, Davanipour Z, Sobel E,Zibulewski J, Schwartz G, Friday G: Multiplesclerosis and childhood infections. Neurology1986;36:1386–1389.

22 Haahr S, Koch-Henricksen N, Moller-LarsenA, Eriksen LS, Andersen HMK: Increased riskof multiple sclerosis after late Epstein-Barr vi-rus infection: A historical prospective study.Mult Scler 1995;1:73–77.

23 Karp CL: Interleukin-12:Amiss in MS. AnnNeurol 1999;45:689–691.

24 Zilber N, Kahana E: Risk factors for multiplesclerosis: A case-control study in Israel. ActaNeurol Scand 1996;94:395–403.