Journal of

J. Neurol. 215, 1--26 (1977)

Neurology © by Springer-Verlag 1977

Original Investigations Geography in Multiple Sclerosis John F. Kurtzke Departments of Neurology and of Community Medicine, Georgetown University School of Medicine, and Neurology Service, Veterans Administration Hospital, 50 Irving Street N.W., Washington, D.C. 20422, USA

Summary. Both mortality and morbidity data indicate quite clearly that multiple sclerosis is a geographically-related disease, and thus MS can be thought of as an acquired environmental (exogenous) illness. High frequency parts of the world for MS are Europe between 65 ° and 45 ° north latitude, northern United States and southern Canada, New Zealand, and southern Australia. These regions are bounded by medium frequency MS regions: in Europe to the north, east, and south; in America for southern U.S.; and the remainder of Australia. Latin America, Asia and Africa are essentially of low frequency from present data. Latitude is not a sufficient criterion: at 40 ° north latitude, MS is high in America, medium in Europe, and low in Asia. All high and medium risk areas therefore are in Europe or European colonies; thus MS is the white man's burden spread from western Europe. Within the U.S., MS is less common among Negroes, Japanese, and possibly Amerindians than in whites regardless of geography. Migration studies among risk areas indicate that migrants keep much of the risk of their birthplace, but also that overall the risk is decreased by high-to-low migration, and probably increased by low-to-high. For the former, it seems that adolescence is the age critical for retention of birthplace risk. Some preliminary data on a possible epidemic of MS are also presented. All the epidemiologic information would be most easily explained if MS were an infectious (?viral) illness with prolonged latency. The proof of this though must come from the laboratory.

Key words: Multiple sclerosis - Geographic distribution - Race - Migration Clustering.

Zusammenfassung. Die Daten tiber Mortalit~it und Vorkommen der Multiplen Sklerose zeigen eindeutig, dal] es sich um eine Krankheit mit einer bestimmten geographischen Verbreitung handelt, und dab die Multiple Sklerose als eine erworbene, (exogene) umweltbedingte Krankheit angesehen werden muB. Gebiete mit hoher Frequenz der Multiplen Sklerose liegen in

2

J.F. Kurtzke Europa zwischen dem 65. und 45. Breitengrad, in den n6rdlichen Vereinigten Staaten, Stidkanada, Neuseeland und Stidaustralien. An diese Regionen grenzen solche mit einer mittleren MS-Frequenz, in Europa ntirdlich, 6stlich und stidlich, in Amerika gilt das fur die si~dlichen Vereinigten Staaten und es gilt fur den Rest von Australien. Lateinamerika, Asien und Afrika zeigen im wesentlichen nach den heutigen Untersuchungen eine niedrige Frequenz. Aber der Breitengrad ist nicht genug: auf dem 40. Breitengrad nt~rdlich ist die Multiple Sklerose in Amerika sehr haufig, mittelhaufig in Europa und niedrig in Asien. Alle Gebiete hohen und mittleren Risikos liegen deswegen in Europa oder in den europ~iischen Kolonien. Die Multiple Sklerose ist also die Biirde des weiBen Mannes, die sich von Westeuropa aus ausgebreitet hat. Innerhalb der Vereinigten Staaten ist Multiple Sklerose seltener bei den Negern und Japanern und wahrscheinlich auch bei den amerikanischen Indianern als bei den WeiBen, unabh~ingig v o n d e r Geographie. Untersuchungen des Einflusses der Wanderung von Risikogegenden aus zeigen, dab die Wanderer das Risiko ihrer Geburtsl~inder mit sich bringen, dab aber im groBen und ganzen das Risiko abnimmt, wenn die Wanderung von risikoreichen zu risikoniedrigen Gebieten erfolgt ist, und dab das Risiko zunimmt, wenn die Wanderung in umgekehrter Richtung erfolgt. Bei der ersteren scheint es, dab die Adoleszenz das kritische Alter ftir die Beibehaltung des Risikos des Geburtslandes darstellt. Es werden vorl~iufige Daten einer m6glichen MS-Epidemie vorgestellt. Alle epidemiologischen Unterlagen werden am leichtesten erklarbar, wenn die Multiple Sklerose eine infektiSse (Virus?) Krankheit mit langer Latenz w~ire. Der Nachweis daftir muB allerdings durch Laboratoriumsuntersuchungen erbracht werden.

Introduction

Multiple sclerosis (MS) is a demyelinating disease of young adults characterized by symptomatology which occurs over a considerable extent of time and often with exacerbations and remissions. There is no known cause for this disorder, and little in the way of effective treatment. Some of the more tantalizing clues to etiology have arisen from considerations of the natural history of MS--i.e., its epidemiology. And among these the most important have concerned the geographic distribution of the disease. It is safe to say that, whatever may ultimately be discovered as to the specific cause of MS, it will have to account for the epidemiologic aspects of the illness. The basic material of epidemiology has to do with geographic distribution, which refers to frequency within specific locales for specific groups. The frequency of disease is best defined by relating the cases to the population from which they are drawn. The three commonly used measures to provide this information are incidence, mortality, and prevalence rates. The incidence or attack rate is the number of new cases of the disease beginning in a given period of time, usually one year, and divided by the population at risk. The mortality or death rate refers to deaths caused by the illness during a given period per unit of population, and again one year is usual. The prevalence rate represents the

Geography in Multiple Sclerosis

3

number of cases present at one time in the population. These rates may refer to the total cases and population, or be subdivided according to specific ages or sex for both cases and population. Age-specific incidence rates for multiple sclerosis in Denmark reveal a sharp increase in adolescence, a plateau in young adult years, and a disappearance of new cases by age 60 (Kurtzke and Hamtoft, 1976). The female rates are notably higher than the male in the younger ages. Age-specific prevalence rates from the same Danish series show a marked peak in the fifth decade, and almost all of the patients are age 20 to 70 at prevalence day (Hyllested, 1956).

Geographic Distribution There is a massive literature on the geographic distribution of multiple sclerosis, and no attempt whatsoever will be made to review either the specific material or its development. I will try to provide a bare minimum on death distributions, and as s u m m a r y a statement as possible on the morbidity data. This though is the epidemiologic information which constitutes the basic "facts" on MS, upon which all hypotheses need be built. Further considerations may be found in reviews prepared for the 1973 Kyoto Symposium on Demyelinating Disease (Kurtzke, in press (a)) and for the forthcoming book by E. J. Field, to whom I am greatly indebted for his permission to reproduce here a considerable portion of my chapter (Kurtzke, in press (b)). The reader should be warned that what follows are my own interpretations of the available material, which eminent authorities have often considered in quite different lights. A m o n g the outstanding workers who should be consulted for a perhaps more balanced view are Acheson (1972), Leibowitz and Alter (1973) and Kurland (1970).

Geography from Death Data A very useful contribution was the paper by Goldberg and Kurland (1962) presenting death rates for a number of neurologic diseases for deaths within the 1951--1958 period. All the rates were age adjusted to the 1950 population of the United States, and referred to diseases coded as the underlying cause of death. These death rates from multiple sclerosis are drawn in Figure 1 from the data of Goldberg and Kurland, but reorganized according to geography. The rates in most of western Europe are in the order of 2 per 100,000 per year or higher, although the northernmost lands are closer to 1 per 100,000, as are Canada, US whites, and New Zealand. Within Europe there seems also a sharp drop to the rates from the Mediterranean basin. South American rates would seem rather low, as are those for US nonwhites (of whom more than 90 per cent are Negro). The Asian and African rates are clearly the lowest recorded.

Geography from Prevalence Data Before World War II, there were almost no proper prevalence surveys for MS. Since the war they have been burgeoning, and now number somewhere near 170. When first reviewed some years ago, it was my impression that the prevalence

4

J.F. Kurtzke

Norway Sweden Finland Denmark Scotland England-Wales No. Ireland

J/// / / / / / / / / / / / / / / / / A ~ / //////////A J// [/~ ~/// / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / A ~

////////A

ireland

W, Germany France Belgium

~//// / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / A [/////Y///////////////////////A

~///~Y / / / / / / / / / / / / / / / / / / / / / / / A ~/// / / / / / / / / / / / / / / / / / / / / / / / / / / A Austria I ~ ///////////////////////A Czechoslovakia I//~/"////////////////////////A Greece V / h'A Italy J~// / / / / / / A Portugal I~///////////////A Israel (Jewish) [//~/"////A MULTIPLE SCLEROSIS Netherlands

Switzerland

U. So. Africa (W) [/~ iceland Canada USA-Total ~/ USA-W USA-NW ~/ Mexico Colombia Uruguay

•/////////////A

Average Annual Age-Adjusted Death Rates, 1951-1958 (from Goldberg & Kurland)

/////A

J//~// A Australia J//'~.~ / / / / / / / J New Zealand>~ [//// / / / / / / / / / / / / / J Philippines Japan h Chile

I

0

1

I

I

I

I

I

0.5 1.0 1.5 2.0 2.5 3.0 Annual Death Rate per 100,000 Population

I

3.5

Excludes Maori

Fig. l. Average annual mortality rates per I00,000 population for multiple sclerosis from various countries within the 1951--1958 period, adjusted for age to the 1950 U. S. population. Redrawn from data of Goldberg and Kurland (1962). From Kurtzke (in press (b))

rates for multiple sclerosis divided the world into three areas comprising high, medium, and low frequency regions (Kurtzke, 1964). This impression was contrary to the prevailing opinion of a rather smooth relationship of MS frequency with geographic latitude. Prevalence by Latitude. The prevalence studies now available for Western Europe are summarized in Figure 2, as a correlation of the rates with north latitude. The numbers refer to specific surveys described elsewhere (Kurtzke, 1975a, b). The solid circles are what I have considered Class A studies, as

Geography in Multiple Sclerosis

~9o F

5

MULTIPLE SCLEROSIS

185 ,J~ 145 ~ _

Prevalence Rates per 100,000 Population for Probable MS

140

~6

115 IfO

WESTERN EUROPE

Z

.r 7

1974

o~ Ioo

~ 9o a. o

~= T

,4

~ 70 ~ 6O

~, 5o Z

40

4o 34

37

w L

4aiz4 5t3

8

30

25

16 I

20 -

.

,o_-. . . . . . . . . . .

.

.

.

.

.

.

.

.

.

.

--;;

I

I

I

I

5

I0

15

20

,

25

:30

35

40

45

50

55

60

65

DEGREES N O R T H L A T I T U D E

Fig. 2. Prevalence rates per 100,000 population for probable multiple sclerosis in Western Europe, correlated with geographic latitude. Numbers identify the specific survey in Kurtzke (1975a, b). Solid circles represent Class A studies, open circles Class B, diamonds Class C, and squares Class E. Vertical bars define 95% confidence intervals of the rates. From Kurtzke (1975 a)

representing proper surveys with well-defined and comparable methodology and diagnostic standards. Those I rated Class B (open circles) are good surveys but have reason(s) why they are not fully comparable to those of Class A. Surveys denoted by diamonds are Class C. These have obvious defects which make them likely to be unreliable, and they should not be taken at face value. Open boxes are Class E surveys, which represent an estimate of the prevalence from the ratios of MS to ALS cases in series from hospitals. Surveys numbered 1--9 are from the United Kingdom, 10--24a from Scandinavia, 25--30 from Germany, 32--40 from France, and 4 1 - - 5 1 a from Italy. Also represented are the Netherlands (No. 31), Spain (40a) and Switzerland (52). It seems to me that MS in Western Europe is in fact distributed according to latitude within two bands: a high frequency zone with prevalence of some 30--80 per 100,000 population (or more) extending from about 43 ° to 65 ° north latitude; and a medium frequency zone with prevalence of some 5--25 per 100,000, and mostly 10-- 15, from about 38 ° to 46 ° . Similar data for Eastern Europe are drawn in

70

MULTIPLE SCLEROSIS

I10 -

Prevalence Rates per 1 0 0 , 0 0 0

Z

_(2

I00 -

Population for Probable MS

90 --

EASTERN EUROPE & ISRAEL 1974

8O

,58

2 0

D.

70

8

60

o. 50 i

~57

61 60 5

~, 4O 53

< 30 56

2O

{>55

IO

0

~::a

I 5

I I0

I 15

I 20

I 25

I ~' 30

71

72 0 { 69{66

I <,7~o 70 I 35 40 45

[ 50

I

I

I

I

55

60

65

70

DEGREESNORTH LATITUDE

Fig. 3. Prevalence rates per 100,000 population for probable multiple sclerosis in Eastern Europe and Israel, correlated with geographic latitude, as in Figure 2. From Kurtzke (1975a)

MULTIPLE SCLEROSIS

22O

Prevalence Rates per 100,000 Population for Probable MS

140 I 130

1

THE AMERICAS

,~9

1974

I

Z 120 I00

0

29o ~. To < w

60

L 40

30 20 I0

........... , o

;- .............. , ,,,o

,,-~,"~ ,o

20

iii

.............................

,

,

,

,

,

,

,

25

30

55

,o

,5

50

55

, ,,4 ~o

,,~ °5

,o

DEGREESNORTHLATITUDE

Fig. 4. Prevalence rates per 100,000 population for probable multiple sclerosis in the Americas, correlated with geographic latitude, as in Figure 2. From Kurtzke (1975a)

Geography in Multiple Sclerosis

7 MULTIPLE SCLEROSIS

I10

Prevalence Rates per 1 0 0 , 0 0 0 I00

Population for Probable MS

9O

AUSTRALIA-NEW ZEALAND 1974

Z

o

r- 80

2

7o

O

8

60 -

~. 50 ~

i

40

-

135

~

134

127

30

__

O

_____L . . . . . .

I I I ................

50

l:iJt

~,33 I

45

40

35

30

25

,a

20

15

,

,

I0

5

0

DEGREESSOUTH LATITUDE

Fig. 5. Prevalence rates per 100,000 population for probable multiple sclerosis in Australia and New Zealand, correlated with geographic latitude, as in Figure 2. From Kurtzke (1975a)

Figure 3. Included therein are studies from Poland (No. 53--57), Czechoslovakia (58--60), R o m a n i a (61--67a), Hungary (68), Jugoslavia (69), Bulgaria (70), Turkey (71--73), Estonia (83), Cherkassy region of the USSR (84), and Israel

(85--87c). Prevalence surveys from the Americas are denoted in Figure 4. Here we see all three risk zones: high frequency from 370--52 ° , medium frequency from 30o--33 °, and low frequency (prevalence less than 5 per 100,000) from 12°--19 ° and from 63 ° - - 6 7 ° north latitude. The coterminous United States and southern Canada are represented by all the surveys from number 88 to 119a, except for numbers 106 (Greenland), 109 (Jamaica), 113 (Alaska), 117 (Netherland Antilles), and 118 (Mexico City). Australia-New Zealand comprise a high frequency zone for 440--34 ° south latitude, and a medium frequency region for 33°--15 ° south (Fig. 5). The recorded rates though which are considered high are toward the lower end of this range. Rates from Asia and the Pacific in the northern hemisphere are all low, except that Hawaii (No. 145, 146) is likely to be in the medium zone (Fig. 6). These study sites extend from 8° to 47 ° north latitude. In the southern hemisphere, with surveys from 30 ° to 6 ° south, all rates from Asia and Africa are also low, except for English-speaking native-born whites of South Africa (No. 156), as seen in Figure 7.

MULTIPLE SCLEROSIS I10

Prevalence Rates per IO0,OUO Population for Probable MS

I00

Z

90

ASIA & AFRICA (NORTHERNHEMISPHERE)

9

~

8o

2

7o

g ~

1974 Pacific Isles

136-146

Mainland Asia 148-152

6o

Africa

153-161

~ 5o ~ 4o <

3O 2O -

-

'~- I

fl47

Di44~ L~lc~o o I --

I0

15

20

[

--

137a'1~ ~ • "c I ~ll-iil,i2 ~143a /SO 25 30 35

~lStl ~ 142e

40

I 2-15o01 -139 45 50

I

I

I

I

55

60

65

70

DEGREES NORTH LATITUDE

Fig. 6. Prevalence rates per 100,000 population for probabJe multiple sclerosis in Asia and Africa (northern hemisphere), correlated with geographic latitude,'as in Figure 2. F r o m Kurtzke (1975 a)

MULTIPLE SCLEROSIS

llO

Prevalence Rates per 100,000 Population for Probable MS

I00 z

o

9o

ASIA & AFRICA (SOUTHERNHEMISPHERE 1974

8

7o

~B

6o

~

5o

z

4o

Pacific Isles

136-146

Mainland Asia 148-152 Africa

153-161

~ 30 2O I0

0 50

I

I

I

45

40

35

]~156 154 155 159 155 57

~1

30

I 25

I ~43 I 20

15

I I0

~~ I 5

DEGREES SOUTH LATITUDE

Fig. 7. Prevalence rates per 100,000 population for probable multiple sclerosis in Asia and Africa (southern hemisphere) correlated with geographic latitude, as in Figure 2. F r o m Kurtzke ( 1975 a)

Geography in Multiple Sclerosis

9

/

0 N

W 0

@

~6

%

o~

~.~ 0

0

\/--

@

~

m

%N = W 0

~c N

s~ o

0

/

¢.~.--

/

0 N 0 ej

! v4

0

@

~

l0

J.F. Kurtzke

o

60N

$

o

o~, •

40N

IO

0

÷÷

--

x

O

x

2014

- -

--

x

0

--

LEGEND 20S

--

HIGH MEO~ LOW

-I good:

6x 40S

--

L poor:

*

o

x

t

0

+

-I

20W

I

0

,

20E

I

40E

60E

80E

Fig. 9. MS frequency by latitude and longitude in Europe and Africa. Ovals and X's represent "good" studies (Class A, B, E), and vertical diamonds and +'s "poor" studies (Class C, %). Solid symbols are high frequency MS survey sites, open symbols are medium frequency, and X's and +'s are low frequency as defined by prevalence rate ranges. Only one study per site is plotted. From Kurtzke (1975b)

Prevalence by Longitude and Latitude. While there seems to be then a reasonably clear separation into three zones of MS, latitude alone appears to be an insufficient criterion both in Europe and America. And in Asia and the Pacific, latitude seems not a factor at all. Figure 8 represents the rates in Europe by both latitude and longitude. Parallels of latitude, with open numbers, run diagonally upward from left to right, while meridians of longitude, with solid numbers, run diagonally downward. The prevalence surveys are designated by the same numbers and symbols as previously. The rates are measured by the height of the vertical bars, with the locus of each study at the foot of each bar. The areas of special interest are the grids between 50° and 40 ° north latitude and 0° and 20 ° east longitude. There is a sharp drop in MS prevalence within these grids, along an arc that extends from about 50° N, 20° E toward about 40 ° N, 0 ° E. To either side of this dividing line, most of the rates are of similar magnitude within their own region: high to the north and distinctly lower to the south. We can therefore simplify the presentation considerably if we accept prevalence rates of 30--80 per 100,000 population (or more) as representing high frequency MS, those of 5--25 as medium frequency, and those under 5 per 100,000 as low frequency MS. We can further group the studies into "good" (Class A, B, E) and "poor" surveys, the latter including not only Class C but also several relative frequencies (percentages) from hospitals within the USSR. In this

Geography in Multiple Sclerosis

11

12

J.F. Kurtzke

60N

•

8

•

II

40N

20N x

LEGEND

[

HiGH MED. LOW good: • o x poor: t ¢ +

40!

160W

14QW

120W

1QOW

80W

60W

Fig. 11. MS frequency by latitude and longitude in the Americas, as in Figure 9. From Kurtzke (1975b)

60N

40N

20N

0

20S

40S 80E

100E

120E

14QE

160E

160E

Fig. 12. MS frequency by latitude and longitude in Asia, as in Figure 9. From Kurtzke (1975b)

Geography in Multiple Sclerosis

13

e~ O

"O

.9 "O

"O

"O

~"O

O O

~-~

14

J.F. Kurtzke

manner we can define the frequency of multiple sclerosis for Europe and Africa in one figure (Fig. 9). More geographic orientation is provided by plotting the surveys for Europe directly on the map (Fig. 10). Here solid squares represent all high frequency surveys regardless of quality, and open (actually shaded) squares represent all medium ones. The triangles reflect rates which may be medium or low, but are likely to be medium. The heavily shaded part of the map defines the high frequency MS zone of Europe, and the dotted portion defines the medium frequency MS band. Within the heavily shaded portion the dashed line may be the actual separation of high from medium risk areas, but the data are inconclusive (Kurtzke, 1975b). In the same fashion as in Figure 9, we can decribe MS distribution in the Americas (Fig. 11) and from Asia (Fig. 12). For the latter, the high-risk "band" is New Zealand, southern Australia--and ocean. To the top left is what may be the eastern limit of the Eurasian medium frequency zone. All other medium frequencies are from Australia, and the remainder of Asia is low. The world-wide distribution of MS, so far as we now know it, is summarized then in Figure 13. Note that latitude alone is really not a good measure of MS frequency. At 40 ° north, for example, MS is high in America, medium in Europe, and low in Asia.

Other Aspects of Geography The division of the world into three risk areas for MS of itself provides few leads as to the nature of this illness, but further considerations of the distribution may take us another step or two. One aspect is the finer structure of the geographic distribution.

Clustering When the entirety of a country is surveyed at a single time by a single team, then it is possible to describe the geographic distribution across the land in some detail. Such surveys have been accomplished for Norway, Denmark, Sweden, Switzerland, Northern Ireland, northern Scotland, The Netherlands, Iceland and Finland. Repeated surveys covering different generations of patients (and doctors) have been accomplished for Norway, Denmark and Switzerland. While the distribution within Northern Ireland was uniform and that within the Netherlands and Iceland rather equivocal, in all other countries surveyed there were very highly significant deviations from homogeneity, and the high-rate areas tended to be contiguous, forming clusters or foci. The differences in the rates between the highest and lowest regions were in the order of six-fold on the average, and thus the variations would seem of biologic as well as statistical significance (Kurtzke, 1966, 1967a). Essentially the same clusters were found when small geographic units rather than the large counties were used as the units for testing (Kurtzke, 1967b). In Denmark the clustering was across middle Jutland on to Fyn, the island next to mid-Jutland; in Switzerland the concentra-

Geography in Multiple Sclerosis

15 240 360

240 320 200

[]

160

120

/%. ~q~xD [] x

80

x

40 X

DX o

0

f

g iX 40

I

80

l

I

120

I

I

[

I

160

New series

Fig. 14. Correlation of the distributions of MS by county between old series and new series of prevalence surveys of three countries, each covering different generations of patients: Denmark (solid circles); Switzerland (X's); Norway (open squares). Each county rate is expressed as the percentage of its respective national (mean) rate. From Kurtzke (1974)

tion was in the northwestern part of the country. The MS distributions were not related to availability of medical facilities, including hospitals, hospital beds and admissions, all physicians, or neural specialists (Kurtzke, 1965b). Not only was there clustering of MS, but in the lands resurveyed a generation apart, there was a very strong correlation between the early and the later distributions, with correlation coefficients of about 0.8 (Kurtzke, 1969, 1974). In Figure 14 the rates by county from each survey are correlated for the three countries with such data: Denmark, Switzerland, and Norway. With this evidence of stability over time, it was thought appropriate to compare studies from neighboring countries to determine whether any broader pattern might be apparent. In this fashion, the high frequency MS areas appeared to describe one single Fennoscandian focus (Kurtzke, 1968, 1974). With the smallest test-units available, this focus extended from the waist and southeastern mountain plains of Norway eastward across the inland lake area of southern Sweden, then across the Bay of Bothnia to southwestern Finland, and then back to Sweden in the region of Ume~ (Fig. 15). This clustering, as well as the broader geographic distributions already considered, mean to me that the occurrence of MS is intrinsically related to

16

J . F . Kurtzke

--70 ....

61g--

,l16°...,J

--6,¢.

6~'.

._-62 °

....

511" - ....

Fig. 15. Distribution of MS in Fennoscandia. Areas significantly above their respective national means (Xa2> 4.0) are in solid black; those high of dubious significance (X] 2.0--4.0) are crosshatched; those insignificantly high (Xa2 < 2.0) are diagonal-lined; those below the national mean are unshaded. Unit boundaries are omitted. Data represent cumulative death rates within 104 small units of Norway (1951--65); prevalence rates from hospital cases within 106 small units of Sweden (1925--34); disability prevalence rates within 20 hospital districts of Finland (1964); and childhood prevalence rates within 23 counties for the national series of Denmark (1949). Fine horizontal shading represents lakes in Sweden and Finland. From Kurtzke (1974)

g e o g r a p h y , a n d t h e r e f o r e t h a t M S is a n a c q u i r e d , e x o g e n o u s , e n v i r o n m e n t a l disease. W h e n a n d w h e r e it m a y be a c q u i r e d m a y be c o n s i d e r e d f r o m m i g r a t i o n studies.

Geography in Multiple Sclerosis

17

Migration R e t u r n i n g to the b r o a d e r d i s t r i b u t i o n of M S , the definition o f regions .differing in risk p r o v i d e s the basis for i n q u i r i n g into the fate o f migrants a m o n g these regions. T a b l e I s u m m a r i z e s m a t e r i a l on prevalence rates (all ages) a m o n g i m m i g r a n t s to a n d f r o m different M S risk areas (Kurtzke, 1972; K u r t z k e a n d K u r l a n d , 1973). T h e rates are those regardless of age at i m m i g r a t i o n a n d o f time o f clinical onset o f M S in reference to m i g r a t i o n . In b r o a d terms, the i m m i g r a n t s do tend to retain the M S risk o f their birthplace. " R i s k " is defined a c c o r d i n g to the s a m e three frequency zones p r e v i o u s l y discussed. The evidence for risk r e t e n t i o n is better for i m m i g r a n t s f r o m high risk a r e a s to low t h a n is the reverse, where there are in fact very sparse data. The evidence is the least impressive for A u s t r a l i a , where the migrants a p p e a r to have a p p r e c i a b l y l o w e r rates t h a n expected for their h o m e l a n d s . This t h o u g h is e x p l i c a b l e by A u s t r a l i a n i m m i g r a t i o n laws dealing with the disabled. F e w o f these M S m i g r a n t s w o u l d be likely to have had clinical o n s e t - - o r at least d i a g n o s i s - - b e f o r e i m m i g r a t i o n . In the S o u t h A f r i c a n study to be discussed below, a b o u t h a l f the E u r o p e a n i m m i g r a n t s were s y m p t o m a t i c before their move. W e r e this to a p p l y to A u s t r a l i a , then the i m m i g r a n t rates there c o u l d p e r h a p s be doubled. The Israeli d a t a s h o u l d be considered f r o m the age a d j u s t e d rates, since their several p o p u l a t i o n s have strikingly different age c o m p o s i t i o n s . It m a y be that the A f r o - A s i a n i m m i g r a n t s (low risk) to Israel ( m e d i u m risk) have an increased f r e q u e n c y over t h a t o f their birthplace, but it is really t o o early to be sure. Table 1. Prevalence rates per 100,000 population for probable Multiple Sclerosis among native-born and immigrants (modified from Kurtzke, 1972) Immigration site according to its own MS risk High S. Australia Medium Perth, W. Aust. Perth, W. Aust. W. Australia Queensland Israel Israel (crude rates) Low South Africa Neth. Antilles Hawaii

Prevalence Rates among Native born

Immigrants from risk areas High Medium Low

38

37

40 a 14 10 9 9b 4 6c 3 5

4

--

87 a 22 31 15 --19b-33

-----

----6b 3

48 59 --35

15 ---

8

----

aage-specific rate, 40--49 years bage-adjusted to 1960 US population c rate of 3 for Afrikaaners and 11 for English-speaking whites

18

J.F. Kurtzke

Table 2. Multiple sclerosis. Average annual age and sex adjusted death rates per 100,000 population (with 95 percent confidence intervals) for residents of California and Washington; recalculated data of Detels et al. (1972) (Kurtzke and Kurland, 1973) Group

US white, born in state US migrants from North a US migrants from Southb

Death Rate California

Washington

0.9 (0.8--1.0) 0.8 (0.7--0.9) 0.4 (0.4--0.5)

1.5 (1.2--1.8) 1.3 (1.1--1.5) 0.5 (0.3--1.0)

a New England, West North Central, East North Central, Pacific, Middle Atlantic census regions bEast South Central, West South Central, South Atlantic census regions

Other data on low to high areas are sparse. Dassel (1972) recorded three instances of MS among immigrants from Indonesia to Holland. Their onsets were at ages 17, 23 and 25 years, and took place respectively 7, 9 and 8 years after their arrival in the Netherlands. They are probably whites of Dutch origin, but neither this nor the population at risk was provided. Regardless, three cases out of what is likely to be a small migrant group looks impressive. We have discovered three instances of exacerbating-remitting MS among a series of some 3400 children who were born in Vietnam of Vietnamese mothers and French fathers, and who came to France under the age of 20. At interview in 1975, 80 percent were age 20--39. The three MS patients each had clinical onset about 15 years after immigration, which for them was under the age often. Their crude prevalence rate for MS was 89 per 100,000, with a 95 percent confidence interval of 18--260. Since the group is still at risk for MS, adjusting this rate for age would be misleading. The age-specific prevalence rate was 169 per 100,000 age 20--29 (confidence interval 35--494). It appears then that these half-Oriental Asians have a significantly higher rate than had they stayed home, but whether it is really as high as expected in France is unknowable at present (Kurtzke and Bui, in preparation). Contrary viewpoints as to migrants from low to high risk have been provided by Detels and by Dean, both of whom claim that individuals born in low-risk areas do n o t increase their risk of MS upon moving. Detels et al. (1972) recorded MS death rates for two states on the west coast of the United States (Table 2). Here the small numbers of southern migrants show no significant difference when they move north, but the northern migrants to California have a significantly lower death rate than those who moved to Washington. Similar findings were presented for the prevalence of MS among migrants to the same states (Detels et al., 1976). Again though the numbers of southern migrants were too few for confidence, in my opinion. Dean et al. (1976), with a comparison of hospitalizations among immigrants and natives of London, recorded many fewer cases in the former than they expected from the latter. However, even if that were valid (and there are questions as to the appropriate expectations), it would not speak to the question

Geography in Multiple Sclerosis

19

Table 3. Migration in multiple sclerosis. Ratiosa of death rates for residence at birth and death in high and low MS risk areas of Norway and United States (Kurtzke et al., 1971) Place of Death

Place of Birth US (1959--61) Norway(1951--65)

High Low

High

Low

High

Low

1.00a 0.87

0.68 0.46

1.00a 0.59

0.57 0.44

a High/High - 1.00

whether there were more or the same number of MS among immigrants as would have occurred in their native lands. There is evidence from other mortality data in the United States and Norway that migrants from low (actually medium) risk areas do increase their risk of MS (Table 3). The ratios for the United States are also the death rates, and the rate of 0.68 for the southern-born who died in northern US was significantly higher than the rate of 0.46 for those born and died in the south (Kurtzke et al., 1971). Unpublished data from a large series of U. S. veterans with MS versus military controls also support the concept of an increased risk of MS among the southernborn who moved north before entry into service (Kurtzke, Beebe and Norman, in preparation). The question of risk of MS in migrant populations gets quite involved, being dependent not only upon a sufficient number of migrants who change their residence from one risk area to another, but also upon their ages at prevalence day, their length of stay in the new land--and their age at migration. Age at Migration. The rates of Tables 2 and 3 indicate quite clearly that birthplace alone is not the critical factor in the risk of MS. Were this the case, the migrants would have had the same rates as for their birthplace. In Table 4 are case-control ratios for residences among a U. S. Army series of "definite + probable MS" (Beebe et al., 1967; Kurtzke et al., 1971). The very clear difference in MS risk (as defined by the ratios) that was present for residence at birth or at entry into service had totally disappeared in the rather short interval between entry and clinical onset. The inference here then is that the critical age in risk of MS is well before onset and well before age 20 or so. From ages of maximal clustering of MS in Denmark and several other features, it was "tentatively concluded that the actual onset of MS appears to take place on the average between the ages of 10 and 15 years, and that there is probably a "latent" or "incubation" period of some 20 years before the onset of clinical symptomatology" (Kurtzke, 1965a, p. 156). Included in this assessment was an ingenious approach to the problem by Poskanzer and his colleagues (Schapira et al., 1963). They considered that, for siblings both of whom had MS, any c o m m o n exposure would likely have occurred while they were living together. The period of cohabitation before onset averaged almost 20 years, and

20

J . F . Kurtzke

Table 4. Residence by tier in United States at birth, at induction and in service, expressed as case-control ratios, for MS patients in the Army in World War II (Kurtzke et al., 1971)

Tier

Time of residence Birth

Northern Middle Southern

Induction

In servicea

1.45 0.86 0.75

1.38 0.86 0.80

1.03 0.99 1.07

Total 1.01 No. of residences 373 P <0.01

1.00 388 <0.01

1.03 578 >0.10

a Residences in service prior to clinical onset only

Table 5. MS prevalence rates, all ages, per 100,000 immigrants in 1960 according to age at immigration (AAI) for northern European immigrants to Republic of South Africa (modified from Kurtzke et al., 1970)

AAI

UK

all

0--14 15--19 20--24 25--29 (20--29) 30--39 40--49 50+

12.8 66.1 31.8 59.4 (45.7) 58.2 57.7 70.5

12.9 81.1 31.3 58.4 (44.9) 52.4 62.4 80.8

N

65

114

the m e a n age o f such c o m m o n e x p o s u r e was 12 o r 13 years o f age, with a m e a n " i n c u b a t i o n " p e r i o d then of some 20 years. The m e a n minimal e x p o s u r e time was 6 years to the first case and 16 years to the second (Kurtzke, 1965 a). A b o u t all these means t h o u g h the range was considerable. A n d the conclusions were b a s e d on natives of high-risk lands. In one study at least, we were able to define a very specific age that was critical for the risk o f M S , a n d this c o n c e r n e d a survey of E u r o p e a n i m m i g r a n t s to S o u t h A f r i c a ( D e a n a n d K u r t z k e , 1971). F o r such i m m i g r a n t s , the M S prevalence rate, a d j u s t e d to a p o p u l a t i o n o f all ages, is p r o v i d e d in Table 5 a c c o r d i n g to age at i m m i g r a t i o n ( K u r t z k e et al., 1970). F o r i m m i g r a t i o n u n d e r age 15 there is the s a m e m e d i u m prevalence rate as for the n a t i v e - b o r n E n g l i s h - s p e a k i n g S o u t h A f r i c a n s ( D e a n , 1967). But for all older age groups, the prevalence is a b o u t w h a t one w o u l d expect f r o m their high risk h o m e l a n d s . I n this series then, age 15 was critical for t h e a c q u i s i t i o n o f M S a m o n g those who came f r o m high risk areas.

R a c e and C o l o r

As recently as 1970, there was for M S " u n c e r t a i n t y as to selection by race" ( K u r l a n d , 1970, p. 75) i n d e p e n d e n t o f geography. W i t h i n the A r m y series n o t e d above, Negroes were a b o u t half as frequent a m o n g the M S as service-wide p r o p o r t i o n s w o u l d suggest, regardless o f b i r t h p l a c e o r i n d u c t i o n residence (Beebe et al., 1967). W e have recently been investigating a m u c h larger series o f all U. S. veterans. These cases we have m a t c h e d with military controls (before they were sick). The full p a p e r is still in p r e p a r a t i o n , b u t in T a b l e 6 are c a s e - c o n t r o l ratios for the m a j o r sex a n d race g r o u p s ( K u r t z k e et al., 1975). There is a clear excess of females a m o n g the whites a n d a clear deficit of b l a c k males, regardless of

Geography in Multiple Sclerosis

21

Table 6. Multiple Sclerosis: Case-control ratios by sex and race according to tier of residence at induction, for veterans of World War II a (Kurtzke, Beebe and Norman, 1975) Sex and Race

Tier of induction Total

North (ratios)

Middle

South

Total

1.00

1.38

1.01

0.53

White female White male Black male Other male

1.91 1.04 0.40 0.29

2.71 1.38 0.550.27b

1.94 1.04 0.50 0.1 lb

0.78 0.57 0.29 0.33

aincludes those who also served in Korean conflict bunstable due to small numbers (case+control <20) Table 7. Multiple Sclerosis: Case-control ratios for "Other" males (neither white nor black) by birthplace and race, for veterans of World War II, and/or Korean Conflict (Kurtzke, Beebe and Norman, 1975) Birthplace and Race

Ratio

Case/control

Coterminous U.S. Amerindian Mexican-Sp. Am. Japanese [All White Male] Mexico, Central Am. Mexican-Sp. Am. Puerto Rico Puerto Rican Hawaii Japanese other China Chinese Philippines, SE Asia Filipino other

0.48 0.38 0.60 0.50 [1.04]

11/23 a ( N 6 / 1 2 S 5 / l l ) b 3/8 (N 3/6 S 0/2) 6/10 (N 1/1 $5/9)

2/4 --

0.00

0/5

0.38

6/16

0.00 0.00

0/l0 0/5

0.00

0/4

0.00 0.00

0/9 0/l

(N 2/4 S 0/0) [ N 1.2 S 0.6]

a One Filipino control included bratios of cases to controls for those born in northern and middle states (N) or in southern states (S)

geography. The ratios for black males are a b o u t the same as we had previously f o u n d with the m u c h smaller sample a n d w i t h o u t matched controls. F u r t h e r detail on the "other" males, i.e. those who were neither white n o r black, is provided in T a b l e 7. The a p p a r e n t deficit for M e x i c a n A m e r i c a n s (6/10) is m o r e likely a reflection of geography t h a n of race. O n the other hand, while the n u m b e r s are too small for confidence, they s u g g e s t a paucity of M S cases a m o n g

22

J.F. Kurtzke

Table 8. Multiple Sclerosis: Case-control ratios according to race and birthplace in selected regions, U. S. veterans of World War II and/or Korean Conflict (Kurtzke, Beebe and Norman, 1975) Region

Mexico, Centr. Am. Puerto Rico Hawaii Japan, Korea China Philippines, SE Asia

Ratio

0.14 0.42 0.06 -0.00 0.00

Case / control Total

White

Black

Other

2/14 14/33 1/16 4/ /4 -/12

1/9 6/14 1/ 1 4/-/ /2

1/2/3 -/ / / -/

/5 6/16 / 15 -/ -/4 / 10

Amerindians (3/8) and American Japanese (2/4)regardless of birthplace. The low ratios for all servicemen born outside the coterminous United States may reflect either race or geography, or both. To explore this further we looked at all foreign-born regardless of race (Table 8). In Latin America including Puerto Rico, there is no evident difference by race, but a significant deficit by geography. In Hawaii the deficit is among Japanese (0/10) and perhaps the Polynesians (0/5), while the single white case has a single control. In the rest of the Orient there is a significant difference for Filipinos as well. Detel's group was unable to find a single case of MS among 12,000 JapaneseAmericans in Seattle, Washington, where the native white prevalence rate was 69 per 100,000. The 95 percent confidence limit on the rate of 0 would be 31 per 100,000. There were 8/84,000 Japanese-Americans born in Los Angeles for a crude prevalence of 10 per 100,000, significantly lower than the native white rate of 22 (Detels et al., 1975; Detels, personal communication). Further, Detels found no cases of MS among 20,000 Japanese immigrants to Seattle or 4000 to Los Angeles. Thus there is growing evidence for racial predilection in MS. Negroes in the United States have less than half the risk of whites, regardless of geography, and this may be true as well for Amerindians. Japanese from Japan, the continental US or Hawaii, and Filipinos from the Philippines are also relatively "immune," as would be the Bantu of South Africa. Therefore MS is the white man's burden. Looking back on the geography, we may see that all high-risk areas for M S - - a n d indeed all medium-risk areas as well--are in Europe or in European colonies (like the United States). It would seem then that MS is the Western European Disease, which has been spread to other parts of the world. "Further work is necessary to see if one can define better the nature and the timing of the spread from Europe to other lands. Unfortunately early information on MS frequency is largely anecdotal. There is some evidence, though, that the distribution of MS in the United States has changed in this century. There is also evidence that the clustering of MS seen in Scandinavia and Switzerland has become less pronounced over time, further suggesting diffusion of the disease (Kurtzke, 1974). There is also the curious difference in risk between the two groups of native whites of South Africa that requires exploration, but that could indicate the establishment of MS as a reasonably common disorder in that land within the past few decades." (Kurtzke, 1975b, p. 153).

Geography in Multiple Sclerosis

23

Comment The prevailing opinion about MS for much of this century to the 1950's had been that MS was some sort of allergic--or hypersensitivity--or a u t o i m m u n e - disease. Thus a bevy of workers have dedicated their lives to EAE (experimental allergic encephalomyelitis). Very few neurologists then thought seriously about an infectious origin, and virologists thought not at all of the disease. The attention of virologists to MS has been the direct result of the epidemiologic studies. Whether this has been a step forward or back remains for the future, but I do believe that the evidence is persuasive that we are dealing with an acquired, exogenous, environmental disease, whose acquisition is long before clinical onset in the normal situation. For natives of high risk areas this acquisition seems to be at adolescence, and one's geographic locale thereafter irrelevant. Explanations for this a m o n g migrants from high to low risk areas fall into two conflicting interpretations, aside from the postulate of some unknown protective factor in the low risk areas. They are: (a) moving to low risk areas before adolescence protects one from the disease because it is thereafter acquired as a subclinical illness in the low-risk areas; and (b) moving thus protects one because he is too young to have acquired it by then in his native land. The former demands the disease agent to be rife where it is clinically absent, and the analogy has been with paralytic vs. non-paralytic poliomyelitis. The latter requires the disease agent to be located in the same place whether it causes clinical or subclinical involvement. It is principally to differentiate between these two major and opposite possibilities that we need definitive information on low to high migration. Were the first (the polio analogy) correct, then migrants from low to high would show n o increase in the risk of MS, since they are already protected by having been exposed from birth. Unless one demands "constant exposure from birth to age X" in order to be protected, any move at age X must be free of increased risk of MS. While still far from being sufficient for full confidence, there is, I think, considerable evidence from the death data and modest evidence from the morbidity work that migrants moving from low to high risk areas d o indeed increase their chances of having MS. Thus the polio analogy seems unlikely to me.

But the primary question remains, if MS is an acquired disease, is it toxin, deficiency, or infection? The epidemiologic evidence would be explained most s i m p l y by an infection. And with the advent of the "slow viruses," a model does exist for an infection persisting for years with little evidence in the neuraxis and with a long incubation. If this is an infection, there should be evidence it is a transmissible illness--in the laboratory or in the populace. The former is not at hand as this is written, though flurries of interest continue to arise as one agent or another is proposed. In the population, one can take transmissibility as o n e explanation for what seems to me the dispersion of this disease from Europe; but this could hardly be thought definitive even if the diaspora is fact. Good evidence for transmissibility would be the description of epidemics of MS. While pockets of cases have been described occasionally, most either do not hold up, or the numbers are so small

24

J . F . Kurtzke

t h a t chance c a n n o t be rejected as their e x p l a n a t i o n . I a m not s p e a k i n g here to the clustering I n o t e d above; this o f i t s e l f - - e v e n if a c c e p t e d - - i s h a r d l y evidence for transmissibility, t h o u g h certainly c o m p a t i b l e with t h a t idea. I t h i n k t h o u g h t h a t we have e n c o u n t e r e d a genuine e p i d e m i c of M S . Hyllested a n d I have been s t u d y i n g M S in the F a e r o e s , that small g r o u p of f o r m e r l y D a n i s h islands lying between N o r w a y a n d Iceland. A p r e l i m i n a r y s u m m a r y dealing with o u r 1974 survey has been p u b l i s h e d ( K u r t z k e and Hyllested, 1975). O u r later visit in 1975, a n d c u r r e n t investigations on five or six m o r e deaths, m a y alter the findings s o m e w h a t , but the m a i n outline still seems valid, I believe. In essence, all of 21 native cases of M S now k n o w n to us as of 1975 to have occurred in the F a e r o e s h a d their clinical onset between 1943 and 1960, save one whose onset was in 1962 after f o u r years in D e n m a r k , a n d one with onset in 1970. The 19 "native" cases o f 1943--60 to me c o m p r i s e an epidemic, with a s u d d e n a p p e a r a n c e and a s u d d e n disappearance o f the disease. As to a likely cause of the epidemic, the only unusual occurrence in the F a e r o e s in recent times t h a t we could discover was their o c c u p a t i o n in W o r l d W a r II by British forces for five years f r o m A p r i l 1940. They n u m b e r e d some 8000 men, or a l m o s t one B r i t o n for every three Faeroese. L o c a t i o n s of the cases a n d the British stations on the islands d u r i n g the w a r were identical for m o s t o f the M S , a n d direct social c o n t a c t with the t r o o p s was a d m i t t e d by most the patients or their survivors. If this holds up, I t h i n k we d o have strong evidence t h a t - - a t least on the F a e r o e s - - M S is a transmissible disease. F u r t h e r s p e c u l a t i o n w o u l d be p r e m a t u r e , b u t I t h i n k this is a very exciting a r e a for o u r c o n t i n u e d investigations.

Acknowledgement. I am especially grateful to Professor E. J. Field of Newcastle, who very kindly allowed me to cite extensively from my chapter in his forthcoming book on Problems in Multiple Sclerosis. I am also grateful both to my co-authors and to other colleagues for their permission to quote data, some of which is as yet unpublished, as well as for their continued help and advice. Among these, my special thanks to Drs. Gilbert Beebe, Roger Detels, Irving Goldberg, and Leonard Kurland.

References Acheson, E. D.: The epidemiology of multiple sclerosis. In: D. McAlpine, C. E. Lumsden, E. D. Acheson, Multiple Sclerosis. A Reappraisal. Second Edition, pp. 3--80. Baltimore: Williams and Wilkins 1972 Beebe, G. W., Kurtzke, J. F., Kurland, L. T., Auth, T. L., Nagler, B.: Studies on the natural history of multiple sclerosis. 3. Epidemiologic analysis of the Army experience in World War I1. Neurology (Minneap.) 17, 1--17 (1967) Dassel, H.: Discussion on the epidemiology of MS. In: E. J. Field, T. M. Bell, and P. R. Carnegie (eds.), Multiple Sclerosis. Progress in Research, pp. 241--242. Amsterdam: North-Holland 1972 Dean, G: Annual incidence, prevalence, and mortality of multiple sclerosis in white SouthAfrican-born and in white immigrants to South Africa. Brit. Med. J. 1967 II, 724--730 Dean, G., Kurtzke, J. F.: On the risk of multiple sclerosis according to age at immigration to South Africa. Brit. Med. J. 1971 III, 725--729 Dean, G., McLoughlin, H., Brady, R., Adelstein, A. M., TaUett-Williams, J.:Multiple sclerosis among immigrants in Greater London. Brit. Med. J. 1976 I, 861--864

Geography in Multiple Sclerosis

25

Detels, R., Brody, J. A., Edgar, A. H.: Multiple sclerosis among American Japanese and Chinese migrants to California and Washington. J. Chron. Dis. 25, 3--10 (1972) Detels, R., Visscher, B., Coulson, A., Malmgren, R.: Evidence for lower susceptibility to multiple sclerosis among Japanese. Neurology (Minneap.) 25, 357 (1975) Detels, R., Visscher, B., Malmgren, R., Dudley, J., Coulson, A. H.: Evidence of protection against multiple sclerosis. Neurology (Minneap.) 26, 357 (1976) Goldberg, I. D., Kurland, L. T.: Mortality in 33 countries from diseases of the nervous system. World Neurol. 3, 444--465 (1962) Hyllested, K.: Disseminated Sclerosis in Denmark. Prevalence and Geographical Distribution, pp. 147 + atlases. Copenhagen: J. J~rgensen 1956 Kurland, L. T.: The epidemiologic characteristics of multiple sclerosis. In: P. J. Vinken, G. W. Bruyn (eds.), Multiple Sclerosis and Other Demyelinating Diseases, pp. 63--84. Amsterdam: North-Holland 1970 Kurtzke, J. F.: General features of the prevalence of multiple sclerosis. J. Indian Med. Prof. 11, 4896--4901, 4895 (1964) Kurtzke, J. F.: On the time of onset in multiple sclerosis. Acta Neurol. Stand. 41, 140--158 (1965a) Kurtzke, J. F.: Medical facilities and the prevalence of multiple sclerosis. Acta Neurol. Scand. 41, 561--579 (1965 b) Kurtzke, J. F.: An evaluation of the geographic distribution of multiple sclerosis. Acta Neurol. Scand. 42 (Suppl. 19), 91--117 (1966) Kurtzke, J. F.: Further considerations on the geographic distribution of multiple sclerosis. Acta Neurol. Scand. 43, 283--298 (1967a) Kurtzke, J. F.: On the fine structure of the distribution of multiple sclerosis. Acta Neurol. Scand. 43, 257--282 (1967b) Kurtzke, J. F.: A Fennoscandian focus of multiple sclerosis. Neurology (Minneap.) 18 (Part 1), 16--20 (1968) Kurtzke, J. F.: Some epidemiologic features compatible with an infectious origin for multiple sclerosis. Add. ad Int. Arch. Allergy 36, 59--81 (1969) Kurtzke, J. F.: Migration and latency in multiple sclerosis. In: E. J. Field, T. M. Bell, P. R. Carnegie (eds.), Multiple Sclerosis. Progress in Research, pp. 208--228. Amsterdam: NorthHolland 1972 Kurtzke, J. F.: Further features of the Fennoscandian focus of multiple sclerosis. Acta Neurol. Scand. 50, 478--502 (1974) Kurtzke, J. F.: A reassessment of the distribution of multiple sclerosis. Part One. Acta Neurol. Scand. 51, 110--136 (1975a) Kurtzke, J. F.: A reassessment of the distribution of multiple sclerosis. Part Two. Acta Neurol. Scand. 51, 137--157 (1975b) Kurtzke, J. F.: An overview of the epidemiology of multiple sclerosis. In: T. Yonezawaet al.(eds.), International Symposium on the Aetiology and Pathogenesis of the DemyelinatingDiseases, Sept., 1973, Kyoto (in press (a)) Kurtzke, J. F.: Multiple sclerosis from an epidemiologic viewpoint. In: E. J. Field (ed.), Problems in Multiple Sclerosis. (in press (b)) Kurtzke, J. F., Beebe. G. W., Norman, J. E., Jr.: Epidemio~ogyof multiple sclerosis in the United States: Preliminary data. Neurology (Minneap.) 25, 356--357 (1975) Kurtzke, J. F., Dean, G., Botha, D. P. J:: A method ofestimatingthe age at immigration of white immigrants to South Africa, with an example of its importance. S. Aft. Med. J. 44, 663--669 (1970) Kurtzke, J. F., Hamtoft, H.: Multiple sclerosis and Hodgkin's disease in Denmark. Acta Neurol. Scand. 53, 358--375 (1976) Kurtzke, J. F., Hyllested, K.: Multiple sclerosis. An epidemic disease in the Faeroes. Trans. Am. Neurol. Assoc. 100, 213--215 (1975) Kurtzke, J. F., Kurland, L. T.: The epidemiology of neurologic disease. In: A. B. Baker, L. H. Baker (eds.), Clinical Neurology Volume 3, pp. 80. Hagerstown, MD: Harper and Row 1973

26

J.F. Kurtzke

Kurtzke, J. F., Kurland, L. T., Goldberg, I. D.: Mortality and migration in multiple sclerosis. Neurology (Minneap,) 21, 1186--1197 (1971) Leibowitz, U., Alter, M.: Multiple Sclerosis. Clues to Its Cause, pp. 373. Amsterdam: NorthHolland 1973 Schapira, L., Poskanzer, D. C., Miller, H.: Familial and conjugal multiple sclerosis. Brain 86, 315--332 (1963)

Received July 15, 1976