Jump to ContentJump to Main Navigation
NeuroepidemiologyFrom principles to practice$

Lorene M. Nelson, Caroline M. Tanner, Stephen Van Den Eeden, and Valarie M. McGuire

Print publication date: 2004

Print ISBN-13: 9780195133790

Published to Oxford Scholarship Online: September 2009

DOI: 10.1093/acprof:oso/9780195133790.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 16 January 2019

Multiple Sclerosis

Multiple Sclerosis

(p.188) 8 Multiple Sclerosis

Lorene M. Nelson

Caroline M. Tanner

Stephen K. Van Den Eeden

Valerie M. McGuire

Oxford University Press

Abstract and Keywords

This chapter provides information on the epidemiology of multiple sclerosis (MS), the most common disabling neurological disease in young adults. It describes the clinical and pathologic features of MS and how these features pose challenges for clinical diagnosis and case definition criteria. Information is provided regarding the descriptive epidemiology of MS, including studies of incidence, prevalence, and temporal trends in MS frequency. Also included is a discussion of the interesting geographical features of the MS distribution, including MS disease clusters, the latitude gradient in disease risk, and migrant studies of individuals who move from high-risk to low-risk regions. Other sections of the chapter cover evidence regarding the infectious etiology of MS, including the important role that Epstein-Barr virus appears to play in disease susceptibility. The role of lifestyle factors is receiving increasing emphasis in MS epidemiologic studies, and evidence is summarized regarding the potential role of cigarette smoking, diet, and hormonal factors.

Keywords:   epidemiology, temporal trends, latitude gradient, migrant studies, disease clusters, infectious etiology, Epstein-Barr virus, cigarette smoking, vitamin D, hormonal factors

Multiple sclerosis (MS) is the most common disabling neurological disease in young adults in North America and Europe. Although treatment may modify the course of the disease, there is no cure. Substantial progress has been made in understanding the pathogenesis of MS, and there is compelling evidence of strong genetic and environmental determinants of risk, but neither has been conclusively identified.

Disease Description

Clinical and Pathological Features

Multiple sclerosis is a chronic inflammatory disease of the central nervous system characterized by large multiple focal areas of demyelination. The lesions, or plaques, are of varying age and are distributed from the cerebrum through the spinal cord. This “scattering in time and place” was recognized by Jean-Martin Charcot (1825–1893) in 1865 as a key feature of the disease. The optic nerves, periventricular areas, corpus callosum, brain stem, and cervical spinal cord are among the commonly affected areas. Microscopically, plaques represent an inflammatory reaction with infiltration of T cells and macrophages, loss of myelin, and reactive glial scar formation. Axons are only relatively spared, and axonal injury in actively demyelinating lesions can lead to the accumulation of irreversible neurological deficit. The symptoms and signs of MS are variable, reflecting the location and severity of the de-myelinating lesions. Common initial symptoms include one or more of the following: weakness in one or more limbs, visual problems (blurring, diplopia), sensory loss or paraesthesiae, and impaired balance. Repeated magnetic resonance image (MRI) scans in MS patients enrolled in clinical trials have revealed that the early course of MS in most patients is more active than suspected from clinical signs, as most lesions seen on MRI scans are asymptomatic (Jacobs et al. 1986; Ormerod et al. 1987; O’Riordan et al. 1998).

The clinical course of MS is extremely variable, both between patients and within a given patient over time (Fig. 8–1). The clinical course of MS has been traditionally described (p.189) as relapsing-remitting, secondary progressive, or primary progressive (Ebers 1998). Not all patients fall neatly into these categories, however, and “transition” forms have been described, as presented schematically in Figure 8–1 (Lublin and Reingold 1996). Most patients experience discrete attacks at the onset of the disease, with complete or partial recovery between attacks (relapsing-remitting MS). About 80% of patients recover from

Figure 8–1 Clinical patterns of multiple sclerosis (MS). Each panel corresponds to different patterns of occurrence of MS symptoms over time. A: Relapsing–remitting with residual deficits after relapses and subsequent progressive time. B: Relapsing-remitting with full recovery followed by relapsing-remitting with residual deficits and subsequent progressive course. C: Chronic progressive with superimposed relapses. D: Chronic progressive from onset. E: Relapsing-remitting with full recovery, long periods of remission, subsequent relapses, and subsequent progressive course. F: Relapsing-remitting with initial full recovery, long periods of remission, and subsequent relapses with residual neurologic deficits. G: Benign MS with relapsing-remitting, with full recovery and subsequent full remission. (Reprinted with permission from Lublin FD, Reingold RC. Defining the clinical course of multiple sclerosis: results of an international survey. Neurology 1996;46:906–911.)

the first attack and remain asymptomatic for variable periods of time (Matthews 1998b). Discrete attacks tend to become less frequent over time and the disease may enter a progressive phase (secondary progressive MS). In about 10% of patients the disease is progressive from onset without recognizable attacks (primary progressive MS); it has been suggested that this form may be a pathologically and etiologically distinct disease.

The spectrum of MS severity ranges from benign disease, with minimal or no disability accumulated even after many years (Thompson et al. 1986), to rapidly fatal massive cerebral demyelination (Marburg’s variant of acute MS). Rare, atypical forms of MS may occur (neuromyelitis optica or Devic’s disease; Balo’s concentric sclerosis) that may or may not have the same etiology as the most typical forms. The MS lesions in a given individual appear to conform to one of four distinct types (Lassmann et al. 2001), suggesting that what we call MS could be in fact the aggregate of four distinct disorders. Although these four types of lesions were defined by histological examination of brain tissue, subsets of MS patients may in the future become identifiable using MRI and magnetic imaging spectroscopy, with important implications for epidemiologic research.

Accuracy of Clinical Diagnosis

The diversity of symptoms and signs and the unpredictable clinical course make the differential diagnosis of MS complex. A large number of conditions can be confused with MS, but most of these are rare, especially in young adults. A detailed account on the differential diagnosis of MS can be found in classical textbooks (Matthews 1998a). As MS is rarely fatal, there is only limited pathological validation of clinical diagnoses. Moreover, autopsy is usually performed selectively and may not provide a good estimate of the predictive value of clinical diagnoses in population-based studies. The largest pathological series is probably that reported by Engell (1988). This author used the Danish MS Registry to identify all patients with a clinical diagnosis of definite MS (p.190) who died during the period of 1965–86 and had a postmortem examination. The diagnosis of MS was confirmed in 485 (94%) of 518 cases, while it was found erroneous in 33, including 9 with brain tumors, 18 with other neurological diseases, and 6 with miscellaneous conditions. Conversely, clinically silent cases of MS discovered unexpectedly at autopsy have also been described.

Case Definition Criteria

No single clinical feature or diagnostic test is sufficient for the diagnosis of MS. Several case definition criteria have been proposed to verify dissemination in time and space of lesions typical of MS and to exclude other explanations of the clinical features. The criteria most commonly used in recent studies are those of the Poser committee (Poser et al. 1983), which were introduced in 1983 (Table 8–1). The four groups defined by these criteria allow some flexibility for investigators in need of definitions with differing degrees of sensitivity and specificity in epidemiological studies. More recently a new classification has been proposed by the international panel on the Diagnosis of Multiple Sclerosis (McDonald et al. 2001), who integrated MRI into the diagnostic scheme, added criteria for the diagnosis of primary progressive disease, and specified that the outcome of a diagnostic evaluation should be one of the following: MS, possible MS, or not MS (see Table 8–2).

The widespread use of MRI has greatly reduced the time from first onset of neurological symptoms to diagnosis and is also likely to decrease the proportion of cases that remain undiagnosed. Gadolinium enhancement identifies putative areas of active inflammatory process (McDonald 1998) and the appearance of new enhancing lesions is considered a clear sign of disease activity. Concern has been expressed, however, that excessive reliance on MRI can lead to overdiagnosis (Kurtzke 1988, Herndon 1994), because current MRI techniques are not specific for demyelination. Until MRI techniques are improved, epidemiologists should classify patients on the basis of both MRI-independent and MRI-dependent criteria: Poser’s criteria to facilitate comparisons with earlier studies and the new recommendations of the International Panel to incorporate the more rigorous MRI results.

Descriptive Epidemiology

Multiple sclerosis is primarily a disease of young adults and is more common in

Table 8–1 Poser et al.’s Diagnostic Criteria for Multiple Sclerosis



Clinical evidence

Paraclinical evidence


A. Clinically defined









B. Laboratory supported definite

















C. Clinically probable












D. Laboratory supported probable




CSF, cerebrospinal fluid; OB/IgG, oligoclonal bands or increased immunoglobulin G (IgG); CDMS, clinically definite MS; LSDMS, laboratory-supported definite MS; CPMS, clinically probable MS; LSPMS, laboratory-supported probable MS.

Reprinted with permission from Poser S, Ritter C, Bauer HJ, et al. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol 1983;13:227–231.


Table 8–2 Diagnostic Criteria for Multiple Sclerosis from the International Panel on Multiple Sclerosis Diagnosis

Dissemination in time


Clinical presentation

Laboratory evidence

Two or more attacks

Objective clinical evidence of two or more lesions

Neither MRI nor CSF negative*

Two or more attacks

Objective clinical evidence of two or more lesions

Dissemination in space demonstrated by MRI or two or more MRI-detected lesions consistent with MS and positive CSF or await further clinical attack implicating a different site

One attack or dissemination in time demonstrated by MRI** or a second clinical attack

Objective clinical evidence of two or more lesions

Dissemination in time demonstrated by MRI** or second clinical attack

Objective clinical evidence of one lesion (monosymptomatic presentation)

Dissemination in space demonstrated by MRI or two or more MRI-detected lesions consistent with MS plus positive CSF

Insidious neurological progression suggestive of MS. Dissemination in time demonstrated by MRI** or continued clinical progression of disability for 1 year

No better clinical explanation

Evidence of inflammation and immune abnormality is essential. Abnormal CSF finding and dissemination in space demonstrated by MRI or abnormal VEP***

If criteria in table are fulfilled, the diagnosis is multiple sclerosis (MS). If the criteria are not completely met, the diagnosis is “possible MS.” If the criteria are fully explored and not met, the diagnosis is “not MS”.

(*) No additional tests are required; however, if tests (magnetic resonance imaging [MRI]; cerebral spinal fluid [CSF]) are undertaken and are negative, extreme caution should be taken before making a diagnosis of MS. Alternative diagnoses must be considered. There must be no better explanation for the clinical picture.

() MRI demonstration of space dissemination must fulfill several criteria listed in Table 1 of McDonald et al. (2001).

() Positive CSF determined by oligoclonal bands detected by established methods; see McDonald et al. (2001).

(**) MRI demonstration of time dissemination must fulfill several criteria listed in Table 2 of McDonald et al. (2001).

(***) Abnormal visual evoked potential (VEP) in addition to MRI demonstration of space dissemination; criteria listed in Table 3 of McDonald et al. (2001).

Source: McDonald WI, Compstona A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the Internal Panel on the Diagnosis of Multiple Sclerosis. Ann Neurol 2001;50:121–127

(p.192) women than men; the F:M ratio is about two in most populations. Incidence peaks late in the third decade of life and the disease is rare in children and older adults. The observation of a latitude gradient in the distribution of MS, marked by increasing rates farther from the equator, has intrigued epidemiologists for decades and provides important etiological clues. The vast literature on this topic must be interpreted with caution, however, as rate estimates are not always comparable between studies. This section highlights the methodological problems encountered in describing the distribution of MS, provides an overview of the worldwide prevalence, and focuses on selected regions and studies that are particularly informative.

Methodological Issues

Estimating Multiple Sclerosis Incidence and Prevalence.

The incidence rates of MS in most high-risk areas range from 1 to 10 cases per 100,000. Not surprisingly, most studies are based on cross-sectional surveys, from which investigators have often derived both prevalence and incidence rates of MS; prospective incidence studies would require the follow-up of large cohorts for long periods of time. An important exception is the MS registry that has been operating in Denmark since 1948 and has provided a valuable source for monitoring trends in MS incidence over time (Koch-Henriksen and Hyllested 1988).

Geographical surveys attempt to identify and confirm all the cases of MS among residents in an area for which demographic data are available at a defined point of time, usually from a census. The first step in case identification is a systematic search of hospitals, outpatient clinics, individual physicians, health insurance organizations, MS societies, and other institutions that provide care for or otherwise may hold information on individuals with MS. The success of this search depends on the probability that an individual with MS is diagnosed and the probability that a diagnosed individual is identified by the search, which is determined in part by the completeness and accessibility of the provider lists (Nelson et al. 1986; Rosati 1994).

The diagnosis of MS is influenced by the standards of medical care, including coverage and quality, the level of MS awareness in the community, and the sensitivity of the diagnostic criteria applied. When these are lacking, the lag time between onset of symptoms and diagnosis is long, the prevalence of undiagnosed cases is high, and MS prevalence is underestimated. Incidence will also be underestimated, as some cases will die before being diagnosed. Surveying small populations can help overcome this problem because intensive efforts can be made to identify and thoroughly review all possible sources of information.

Example 8–1 An example is a study of the prevalence of MS in Northern Colorado (Nelson et al., 1986) where the investigators used a variety of methods, including the direct review of records from all neurologists, to strive for near-complete case ascertainment of MS patients in the region. The estimates of MS prevalence from this study were approximately 40% higher than prevalence estimates for the same latitude derived from a survey of randomly selected hospitals and physicians in the United States (Baum and Rothschild 1981). Unlike the Colorado study, this national study did not survey MS service organizations nor audit the case records of practicing neurologists. The estimates from the national survey can be considered an average rate whereas the Colorado study yielded an observed rate for a specific period of time.

Because of improvements in medical care, increased MS awareness, and widespread use of MRI, the lag time between onset of symptoms and diagnosis is decreasing. Faster diagnoses may spuriously increase incidence rates for recent time periods, so some investigators have relied exclusively on clinical criteria to define MS cases when studying time trends (Wynn et al. 1990).


Example 8–2 The medical records should also be searched for conditions that present with clinical features similar to those of MS and records that may include misclassified MS cases. For example, in a study in Israel, the investigators reviewed the medical records of all hospitals, referral clinics, and chronic care facilities for the period 1955 through 1959 for the following diagnoses: MS, primary lateral sclerosis, nontraumatic paraplegia, optic or retrobulbar neuritis, cerebellar ataxia, and myelopathy (Alter et al. 1962). After review of the medical records, they accepted 193 cases as probable MS and 89 cases as possible MS, of which 11 and 25, respectively, were found under different diagnostic labels.

Capture–recapture methods have been used in some surveys to assess the completeness of case ascertainment (Forbes and Swingler 1999). These methods should be applied with caution, as the assumptions required to estimate prevalence are often not satisfied (Hook and Regal 1995); however, a comparison of the proportion of cases identified by different sources can provide some sense of the overall level of under-ascertainment.

The second critical step in case ascertainment is the diagnostic confirmation of the cases detected in the first phase of the survey. Methods have ranged from accepting the hospital or physician diagnoses without further documentation (Svenson et al. 1994) to one or more of the investigators conducting the clinical examination themselves (Alter et al. 1962; Hammond et al. 1988). The impact on the results is important. In the study conducted in the early 1960s in Israel (Alter et al. 1962), for example, the authors personally examined 520 presumed MS cases identified from a medical record review. Of these, only 193 (37%) were accepted as probable and 89 (17%) as possible MS. Even investigations that rely on direct examination of all suspected cases of MS may not be comparable if different diagnostic criteria are used (see Compston 1998a for a comparison of U.K. surveys that used different diagnostic criteria).

Finally, it is important to estimate accurately the denominators of the prevalence and incidence rates and to age-adjust to a common standard for comparison. All of the rates that we will directly compare in this chapter are age–adjusted, unless otherwise stated.

Fortunately, the difficulties of comparing incidence or prevalence rates from different studies were recognized early in the history of MS epidemiology, and efforts have been made to survey different geographical areas with similar methods and diagnostic criteria. Examples include the investigations that established the existence of a strong latitude gradient in Australia (Hammond et al. 1988) and New Zealand (Skegg et al. 1987) and the 10-fold difference in MS prevalence in Sicily (Dean et al. 1979) as compared with that in Malta (Vassallo et al. 1979).

Age at onset.

The MS literature is replete with comparisons of age at onset between different populations or groups within a population. Almost universally these are crude comparisons, sometimes of groups with very different age structures. These comparisons are meaningless and should be ignored, except when the age structure of the compared populations is similar or can otherwise be taken into account.

Temporal changes in multiple sclerosis frequency.

Time trends in the incidence and prevalence of MS can be estimated in certain populations that have been repeatedly and consistently surveyed. Information on time trends in MS incidence in the United States is sparse. The best information, from the Rochester Epidemiology Project, a longitudinal study in Olmsted County, Minnesota, shows that the incidence of MS increased in this century from 1.2 per 100,000 (95% CI 0.0–2.5) in 1905–1914 to 6.5/100,000 (5.5 to 7.5) in 1975–84 (Wynn et al. 1990). The veracity of this increase is supported by the fact that (p.194) 92% of the cases had clinically definite MS; MRI results, which were not available for the earlier time period, were disregarded. In more recent years, surveys in some areas of southern Europe have provided much higher estimates of prevalence than studies from earlier time periods (Rosati 1994). In the United Kingdom, prevalence in the southern part of England has been steadily rising but has remained stable in northeast Scotland (Compston 1998a). Whether these findings represent true changes or simply reflect better ascertainment of cases is a matter of controversy (Rosati 1994; Compston 1998a).

Scandinavia is one of the most intensively studied regions for MS frequency because of the region’s stable and relatively homogeneous population, high standards of medical care, free universal access to health services, and high level of education and MS awareness, as well as the existence of some MS registries. In spite of these advantages and decades of repeated surveys, the findings regarding time trends in MS prevalence are still controversial. The emerging pattern is that of inconsistent fluctuations across the region (Koch-Henriksen 1995). Incidence rates have apparently increased between the 1950s and 1980s in Norway (Grønning et al. 1991; Midgard et al. 1991, 1996) but have decreased or remained stable over the same period in Gotenburg, Sweden (Svenningsson et al. 1990). In Denmark, the incidence was highest in 1950, declined to a minimum in 1967, and then increased again, but without reaching the 1950 level (Koch-Henriksen 1995). The existence of a well-established registry in Denmark and the application of uniform diagnostic criteria make the Danish fluctuations difficult to explain by bias. The investigators suggested that they reflect changes in environmental factors, but skepticism about this explanation remains (Compston 1998a).

Geography and Multiple Sclerosis

The interest in the geographic distribution of MS is motivated by the hope of finding etiological clues. In this context, the main question is whether there are regional variations in the incidence of MS that cannot be explained by genetic differences and may thus provide a clue to an environmental cause.

Because of the scarcity of incidence data, worldwide maps of MS distribution are based mostly on prevalence (Rosati 2001; Pugliatti et al. 2002). Early studies of MS prevalence established a latitude gradient, with prevalence increasing with increasing distance from the equator in both hemispheres. The few studies conducted close to the equator consistently reported a low prevalence of MS (subtropical part of Australia, South Africa, and Israel) (Kurtzke 1995). Kurtzke (1995) created a simplified map based on areas of high (over 30/100,000), medium (5–29/100,000), and low MS prevalence (<5/100,000). In spite of the methodological difficulties in comparing prevalence of MS in different countries, few epidemiologists question the existence of a latitude gradient, as this is quite strong, it has been consistently found in independent studies in the United States, and it was shown in Australia and New Zealand after repeated rigorous surveys conducted with similar methods.

In the United States, a north–south gradient in MS distribution was noted in the 1920s and was originally attributed to the higher proportion of individuals of Scandinavian ancestry in the northern states (Davenport 1921). Since then, several studies have independently demonstrated that MS incidence and prevalence have been substantially higher in the northern part of the United States (Visscher et al., 1977; Kurtzke et al. 1979; Baum and Rothschild 1981; Hernán et al. 1999) and in Canada (Kurland et al. 1952; Stazio et al. 1967) than in the southern United States (Fig. 8–2a). Although any one of these studies is subject to scrutiny because of methodological difficulties in comparing MS rates among different regions, taken together, these studies provide solid (p.195)

                      Multiple Sclerosis

Figure 8–2 A: Multiple sclerosis (MS) prevalence and latitude for surveys conducted between 1965 and 1985, north of the equator (North America). a: MS prevalence for year 1970, Los Angeles, CA (Visscher et al. 1977). b: MS prevalence for year 1965, Rochester, MN (Kurland et al. 1965). c: MS prevalence for year 1976, for regions above the 37th parallel (Baum et al. 1981). d: MS prevalence for year 1970, Seattle, WA (Visscher et al. 1977). e: MS prevalence for year 1982, northern Colorado (Nelson et al. 1986). B: Multiple sclerosis prevalence and latitude for surveys conducted between 1981 and 1983 that use similar methodology, south of the equator (Australia, New Zealand). Surveys: (a) tropical Queensland, Australia; (b) subtropical Queensland, Australia; (c) Perth, Australia; (d) Newcastle, Australia (e) Adelaide, Australia; (f) Waskato, New Zealand; (g) Wellington, New Zealand (h) Hobart, Australia; (i) Otago Southland, New Zealand. (Reprinted with permission from Miller DH, Hammond SR, McCleod JG, et al. Multiple sclerosis in Australia and New Zealand: are the determinants genetic or environmental? J Neurol Psychiatry 1990;53:903–905.)

evidence of an MS latitude gradient in North America.

Example 8–3 The first study that provided strong evidence for a gradient in the incidence of MS was the landmark investigation of U.S. Army veterans (Kurtzke et al. 1979). The study population comprised 5305 veterans with a diagnosis of MS occurring before 1969 and an equal number of matched controls. Since veterans with MS occurring during service or within 7 years from discharge were eligible for recompensation, the degree of case ascertainment was likely to be very high. A review of the clinical records of a random sample of the cases found that 96% met the criteria for definite MS (Kurtzke et al. 1979). The relative incidence of MS was 2.7 among veterans who entered the service in the northern United States (states generally north of 41°–42° north latitude) as compared with those who entered the service in the southern United States (states lying south of 37° north latitude); incidence for the middle states was in between.

Studies from the Southern Hemisphere (New Zealand, Australia) have been methodologically similar and close in time, providing clarity about the nature of the geographic gradient and the possible role of genetic factors in explaining the gradient. A combined analysis of the data from Australia and New Zealand revealed a sevenfold gradient in MS prevalence between Queensland in tropical Australia (12.4 per 100,000) and Otago in southern New Zealand (81.7 per 100,000) (Miller et al. 1990). It may be noted that although (p.196) these data display a strong linear correlation with latitude (Fig. 8–2b), they are more consistent with low, intermediate, and high prevalence zones.

When considered together, the studies from both hemispheres support the interpretation that the frequency of MS increases as distance from the equator increases. Epidemiologists have been searching for factors that might explain this striking geographic distribution. We focus here on selected studies that allow a separation of genetic from environmental factors, and include information from migrant studies as they are critical for this purpose.

Is the Gradient an Effect of Genetics?

Many studies have examined the role of ethnicity or ancestral background as a surrogate measure of a possible genetic explanation for the geographic gradient. Data from the U.S. Army veterans study (Kurtzke et al. 1979) were used to assess the role of genetic factors in the MS latitude gradient in the United States (Bulman and Ebers 1992). Strong correlations were found between the case/control ratios and the proportion of the population in each state with Scandinavian or northern European ancestry, and the authors suggested that genetic susceptibility may explain the distribution of the disease in the United States. An independent analysis of the same data, however, challenged this conclusion, showing that latitude explained 66.5% of the variance in MS risk, whereas Swedish ancestry, the strongest predictor after latitude, explained only 9% (Page et al. 1993). The contributions of other ancestries were small (3.9% for French, positive, and 2.3% for Dutch, negative correlation; other ancestries not significant). These results, although consistent with a contribution of genes to the distribution of MS in the United States, strongly suggest a predominance of environmental factors.

Genetic explanations of geographical variations have also been proposed for New Zealand and Australia. For example, New Zealand has a higher proportion of people of Scottish ancestry, who are believed to have a genetically determined higher risk, in the south (Miller et al. 1990). Subsequent studies, however, revealed that variations in the proportion of people with Scottish ancestry had only a modest impact on MS rates in New Zealand, and that variations in MS frequency in Australia and New Zealand cannot be explained by genetic factors (Swingler and Compston 1986; Hammond et al. 1988; Robertson and Compston 1995).

Migrant Studies

Early investigators thought that migrant studies might help elucidate the contribution of geography to MS risk, primarily by tracking migrant populations who moved from high-risk regions to low-risk regions. The results of these studies and related methodological issues are thoroughly reviewed elsewhere (Gale and Martyn 1995). Migrants often differ in socioeconomic status and education from non-migrants, may have a different genetic constitution, and are usually less likely to have a chronic illness (that is both a disincentive to migrate and at times a barrier to immigration as the host countries may impose restrictions on people with disabilities). There is often scarcity of data on the frequency of MS in the country of origin, availability of medical care may differ from the host country of origin and that of destination, and migrants may differ between population in their use of health services and probability of being diagnosed (Gale and Martyn 1995). Denominators are often uncertain, as up-to-date figures may not be available and illegal immigrants are not included in the statistics. Despite these limitations, migrant studies have provided compelling evidence on the role of environmental factors and have shown convincingly that migration from a high- to a low-risk zone is associated with a reduction in risk. Two points that remain unclear are whether this reduction in risk benefits only those who migrate in childhood and whether migration in the opposite direction (low risk to (p.197) high risk) is associated with an increased risk.

Migration from High- to Low-Risk Areas.

Several studies have considered changes in MS risk in immigrants who migrate from European countries to lower-risk areas of Israel, South Africa, and Australia, and in people who migrate from northern to southern areas of the United States (Table 8–3a). In general, these studies have supported the hypothesis that migration from a high- to a low-risk area in childhood decreases the risk of MS, but they are equivocal as to whether migration as an adult confers any benefit. This pattern is consistent with an environmental risk factor, possibly one that operates in childhood.

Because of its large immigrant population and good access to health care, Israel provides an interesting setting for the investigation of migration and MS. A nationwide survey of MS completed in 1961 revealed that the prevalence of MS among European immigrants to Israel was over six times higher than that among Afro-Asian immigrants (Alter et al. 1962) and that the incidence rate ratios comparing European with Afro-Asian immigrants were 1.7 for those who immigrated before age 15, 8.9 for those who immigrated between age 15 and 29 years, and 5.4 for those who immigrated between age 30 and 44 years (Alter et al. 1966, 1978). Despite the small number of cases among immigrants below age 15 (four among Europeans and four among Afro-Asians), this result adds to the evidence that immigration in childhood from high- to low-risk areas reduces the risk of MS. The effects of migration in adulthood cannot be determined from these data, as comparable rates in the country of origin of the migrants are not available.

One of the most influential studies concerning the importance of age at migration was conducted in South Africa (Dean 1967). The estimated age-adjusted incidence rates of MS were 0.4 per 100,000 among South African–born whites and 2.8 among immigrants from the United Kingdom, which is similar to that of their country of origin. The observation that people of similar British stock had quite different rates of disease depending on their place of birth provides further evidence of the importance of environmental factors. Noting that white children born in South Africa are more likely to be exposed early in life to several infections than those born in the United Kingdom, Dean interpreted these data as supportive of the “poliomyelitis hypothesis,” i.e., a virus that increases the risk of MS if acquired after infancy or in early childhood (Poskanzer et al. 1976). Consistent with this interpretation was the higher rate of poliomyelitis among immigrant children than that in South-African born whites. This survey was updated in 1968 to study the effects of age at immigration on risk of MS (Kurtzke et al. 1970). The number of cases of MS among individuals who were 15 or younger at migration was about 3 to 4 times lower than expected under the hypothesis that age at immigration was irrelevant. In spite of some uncertainty concerning the indirect calculations of the expected number of cases, these results support a stronger effect of migration in childhood than later in life. The number of cases was too small to determine whether there is a specific age at which this benefit is lost. Most importantly, since it is not known what the rates of MS among these immigrants would have been if they had remained in their countries of origin, the overall effect of migration on risk of MS cannot be reliably estimated from these data.

Compelling evidence for the predominantly environmental origin of the MS gradient in the United States is provided by the U.S. Army veterans study. The case/control ratio for veterans born in the north tier decreased from 1.48 if they remained in the northern tier to 0.74 (0.47 to 1.14) if they entered service in the southern tier (Kurtzke et al. 1985). One U.S. study (Detels et al. 1978) found that the benefit of migration from a high- to a low-risk area declined with age, but was still present for migration between 15 and 19 years of age. A (p.198)

Table 8–3a Migration Studies of Multiple Sclerosis Frequency: Migration from High-risk to Low-risk Areas


Migration route


Age at migration effects

Dean 1967; Kurtzke et al. 1970

United Kingdom to South Africa

Sevenfold higher MS prevalence in U.K. immigrants than in South African-born whites

Three- to fourfold lower MS prevalence than expected for U.K. immigrants who were ≤15 years old at migration

Alter et al. 1962 1966; 1978

Europe to Israel

Sixfold higher MS prevalence in European immigrants to Israel than in Afro-Asian immigrants to Israel

Only 1.7-fold higher in European immigrants who were ≤15 years old at migration

Detles et al. 1978

Northern United States (42° latitude) to King and Pierce Counties, WA or to Los Angeles, CA

Lower MS prevalence for those moving from northern U.S. to Los Angeles, CA, than for those moving to King and Pierce, WA

Fivefold lower MS prevalence for migration to Los Angeles, CA, for those ≤9 years old at migration; threefold lower for ages 10–14; twofold lower for ages 15–19 years

Kurtzke et al. 1985

Northern to Southern United States

Decrease in risk of MS among white U.S. veterans born in the north and entering into active duty in the south


Hammond et al. 2000a

United Kingdom to Australia

Risk of MS was low in U.K. immigrants who migrated to an area of Australia where MS is rare

None—reduced MS risk seen at all ages

(p.199) study from Australia (Hammond et al. 2000a) found both adults and children who migrated from the United Kingdom to Australia were protected, suggesting an environmental factor operating also during adult life and not only in childhood.

Migration from Low-to High-Risk Areas.

The effect of migration from low- to high-risk areas is controversial. In general, studies suggest that migrants tend to retain their low-risk level, but results are mixed. Some studies suggest that subsequent generations have increased risk (Table 8–3b).

In the U.S. veterans study, migration from the southern to northern United States was associated with a nonsignificant increase in risk of MS (Kurtzke et al. 1985); however, the period of follow-up may have been too short (9 to 10 years) for a complete expression of the increased risk associated with migration (Kurtzke et al. 1998). In the same study, case/control ratios increased from 1.0 to 1.4 (1.1 to 1.8) among those veterans who were born in the middle tier and entered service in the north tier. Investigations conducted among immigrants to the United Kingdom suggested that migration from low- to high-risk areas does not increase the risk of MS (Dean et al. 1976, 1977). The interpretation of these data is difficult, however, because most immigrants from low-risk areas had been in the United Kingdom for less than 5 years, which may not have been sufficient time for the effect of migration to become manifest (Kurtzke 1976). A later study in the United Kingdom confirmed the scarcity of deaths attributed to MS among Asian immigrants (6 vs. 82 expected) or immigrants born in the West Indies (12 vs. 59 expected) (Elian and Dean 1993). The data above could of course reflect a lower genetic susceptibility to MS among Asians and other immigrants. However, U.K.-born children of these immigrants were later found to have rates of MS substantially higher than those of their parents and similar to those in the general population in England and Ireland, a finding confirming their susceptibility to MS (Elian and Dean 1987; Elian et al. 1990).

Similarly, the low rates of MS in Asian and African immigrants to Israel have been cited to support the hypothesis that the move from low- to high-risk areas is not associated with an increased risk of MS (Gale and Martyn 1995). Risk does appear to increase in ensuing generations, however; prevalence of MS among Israelis whose fathers were born in Africa or Asia was 1.4 to 1.8 times higher than among African and Asian immigrants (Gale and Martyn 1995).

With the exception of the U.S. veterans study, none of the investigations described in this section provides a direct comparison of risk in migrants from low- to high-risk areas with that for people of the same ancestral origin who remained in their country of origin, although they do suggest that any increase in risk of MS caused by migration from a low- to a high-risk area is modest compared to the high risk that migrants would have experienced had they been born in the host country. Even this conclusion is uncertain because comparisons are made between different generations that may have different risks of MS independently from migration.

Multiple Sclerosis Clusters

The Faroes are a group of islands in the North Atlantic Ocean that have been a semi-independent unit of the Kingdom of Denmark since 1948 (Kurtzke and Heltberg 2001). Intensive efforts to ascertain all cases of MS among Faroese from 1900 to 1983 started in the 1940s and 32 cases of MS were identified among native resident Faroese (Fog and Kyllested 1966, Kurtzke and Hyllested 1979, Kurtzke 1986). The absence of cases with onset before 1943, and the clustering of cases with onset between 1943 and 1949, was interpreted as supporting the occurrence of an epidemic of MS. The observation that the beginning of the epidemic started soon after the British occupation of the islands (April 1940 to 1944) and the strong and highly (p.200)

Table 8–3b. Migration Studies of Multiple Sclerosis Frequency: Migration from Low-risk to High-risk Areas


Migration route


Next generation effects

Kurtzke et al. 1985

Southern to northern United States

Nonsignificant increase in risk of MS


Kurtzke et al. 1998

North Africa to France

1.5-fold higher MS prevalence among North African migrants than among general French population


Kahana et al. 1994; Gale and Martyn 1995

Asian and African immigrants to Israel

Maintained low MS risk of country of origin

1.4- to 1.8-fold higher risk of MS in second-generation Afro-Asian immigrants than in first-generation immigrants

Dean et al. 1976, 1977

India, Pakistan, and the West Indies to Greater London

Low rates of hospital admission for MS among immigrants; rates high for immigrants from high-risk areas


Elian and Dean 1987, 1993

Asia or West Indies to United Kingdom

Fewer MS deaths than expected in immigrants

U.K.-born children of immigrants had higher rates of MS than parents and similar rates to native English and Irish

N/A, not applicable.

(p.201) significant association between troop location and parishes of residence (Kurtzke and Heltberg 2001) led the authors to conclude that the epidemic was caused by some agent introduced by the British troops. On the basis of further analyses of the time course of the epidemic, it was concluded that this agent was a widespread, specific, persistent infection, probably asymptomatic (Kurtzke and Heltberg 2001). Critics questioned whether an epidemic had occurred at all (Dean 1988; Poser and Hibberd 1988; Poser et al. 1988), noting that the date of onset of illness is irrelevant, as the disease is supposedly acquired between the ages of 5 and 15 years (Poser et al. 1988). In our view, this is a rather weak objection, because the proposed narrow range of susceptibility is hypothetical; further, the introduction of a new infectious agent in a nonimmune population may result in a different pattern of age susceptibility than what is typical in populations where the infection is endemic. Other criticisms, including suggestions that case exclusions were arbitrary (Poser et al. 1988) and that case ascertainment was incomplete, seem unlikely to account for the sudden excess of cases seen in the 1940s, and the claim of an epidemic occurrence of MS in the Faroes seems well founded. The detailed interpretation offered by the investigators remains more speculative, but overall the Faroese experience supports the play of environmental factors in MS incidence.

Genetic Epidemiology

Familial Studies

Population-based twin studies of MS have consistently shown a higher concordance rate among monozygotic (MZ) than dizygotic (DZ) pairs. As shown in Table 8–4, these rates ranged from 26% to 31% for MZ pairs, and from 0% to 5% for DZ pairs (Bobowick et al. 1978; Heltberg and Holm 1982; Ebers et al. 1986; Kinnunen et al. 1988; Sadovnick et al. 1993; Mumford et al. 1994). The risk in non-twin siblings of an MS proband was similar to that of DZ twins. These results clearly indicate the importance of genetic factors in determining the susceptibility to MS, as the degree to which MZ and DZ pairs share the same environment should be similar, and DZ twins are as genetically similar to each other as non-twin siblings. Further, the large difference in risk between MZ and DZ twins and other siblings provides strong evidence against the existence of a single disease susceptibility locus, as this would imply approximately a twofold gradient in risk between MZ and DZ twins (Risch 1990). Age-adjusted recurrence risks in first-degree relatives of MS probands are elevated and appear to vary by the nature of the relationship (e.g., child or sibling) (Sadovnick et al. 1993; Robertson et al. 1996; Carton et al. 1997). Second- and third-degree relatives also had substantially higher risk of MS than expected from general

Table 8–4 Percent Concordance for Multiple Sclerosis Among Monozygotic and Dizygotic Twins





% Concordance


% Concordance

Bobowick et al. (1978)





Heltberg and Holm (1982)





Ebers et al. (1986)





Kinnunen et al. (1988)





Sadovnick et al. (1993)





Sadovnick et al. (1993)*





Mumford et al. (1994)





(*) Adapted from Ebers et al. 1986, with permission; one case from original series did not have MS and one dizygotic pair became concordant.

() Same-sex twins.

(p.202) population rates (Robertson et al. 1996). Since these relatives are less likely to share a common environment, this further supports the importance of genetic factors.

As members of the same family share a common environment and lifestyle, non-genetic factors could contribute to familial aggregation.

Example 8–4 Studies of half-siblings, children adopted by individuals with MS, and parents who adopted a child who developed MS were conducted to address this possibility. From a population-based sample of 15,000 individuals with MS in Canada (Ebers et al. 1995), investigators selected Caucasian adoptees with MS who were adopted as infants and determined the occurrence of MS among their nonbiologic relatives, including 470 parents, 345 siblings, and 386 children. In each group, the age-adjusted risk was similar to that expected in the general population and significantly less than predicted from the biological risk data. These results provide strong evidence for a genetic cause for familial aggregation. In a separate study, the risk of MS was compared between full-sibs and half-sibs of the cases. The age-adjusted MS risk was 1.32% in 1839 half-sibs and 3.46% in 1395 full-sibs (p < 0.001). There were no significant differences in risk for half-sibs raised together or apart from the index case (1.17% vs. 1.47%) or for maternal versus paternal half-sibs (1.42% vs. 1.19%) (Sadovnick et al. 1996). This last finding, and the fact that children of an affected father have similar risk to that of children of an affected mother, suggests that the impact of maternal effects, such as imprinting, mitochondrial inheritance, intrauterine factors, or breast feeding, must be modest (Sadovnick et al. 1997). Further analyses revealed a low prevalence of conjugal MS and a high risk of MS in offspring of MS parents (Ebers et al. 2000), comparable to the risk of MZ twins of affected individuals. This high rate is consistent with a role of recessive alleles that are shared among unrelated individuals with the disease. The small number of cases and concerns over the appropriateness of age adjustment (which does not take into account that children with both affected parents may have an earlier age at MS onset than average) dictate some caution in interpreting this provocative result (Ebers et al. 2000).

A high risk of MS among children of conjugal pairs was also suggested in a U.K. study (Robertson et al. 1997). Five of 86 offspring from conjugal pairs in which both parents had MS met the criteria for clinically definite MS, and a further 5 had either characteristic MRI abnormalities or clinical symptoms consistent with demyelination, but did not meet criteria for clinically definite disease.

In summary, the results of familial studies indicate that aggregation of MS within families is due to shared genes rather than environment, that more than one gene contributes to MS susceptibility, and that there is little or no role of maternal factors in the inheritance of MS.

Candidate Genes

Whereas evidence for a strong genetic component in MS is undisputable, progress in identifying the responsible genes has been disappointing. If inheritance of MS susceptibility is polygenic, each of the involved genes may by itself only provide a modest increase in risk and could be difficult to identify. A detailed review of the genetic epidemiology of MS has been published (Compston 1998b). Only selected findings will be discussed below.

HLA Region.

The only region that appears certain to contain a gene that confers susceptibility to MS is the major histocompatibility complex (MHC), or HLA encoded on chromosome 6q21.1–21.3 (Compston 1998b). Within this region, the most consistent association is with the class II alleles defined in serological tests as HLA-DR15 and DQ6 (DR15 is a split of the serological specificity previously identified as DR2). The serological specificity DR15 is the expression of the genotype DRB1*1501, DRB5*0101 (DRB1 and (p.203) DRB5 are two different loci both encoding the β chain of the HLA class II molecule in the DR region; the numbers after the asterisk denote specific alleles), whereas DQ6 is the expression of DQA1*0102, DQBl*0602 (DQA1 is the locus encoding the a chain, and DQB1 is the locus encoding the β chain of the class II molecule in the DQ region). Because of strong linkage disequilibrium, the individual contributions of DR and DQ polymorphisms have been difficult to separate. Previously reported associations with class I alleles (A3 and B7) appear to be due to linkage disequilibrium with DR15 and DQ6. The association with DR15 has been found in most but not all populations. A notable exception is Sardinia, where the strongest association is with DR4. In a recent case–control study of 159 patients with definite relapsing-remitting MS and 273 control subjects, investigators reported an association between an allele in a microsatellite marker telomeric to the HLA class II region and the risk of developing relapsing-remitting MS (OR = 2.0, 95% CI 1.2–3.1; p = 0.004). This allele is encoded within an ancestral haplotype that is highly linked to HLA-DR3. The joint effect of this ancestral haplotype and HLA-DR2 on the risk of developing relapsing-remitting MS was 8.7 (95% CI 2.7–29; p < 0.0001). This result suggests that within the HLA region, other loci besides HLA-DR2 haplotype modulate susceptibility for relapsing-remitting MS (de Jong et al. 2002). Kalman and Lublin (1999) and Giordano et al. (2002) have recently reviewed the associations of HLA and MS. Overall, the associations with DR15 or other HLA class II alleles are too weak to explain most of the genetic variability of MS, which suggests that other important loci may exist.

Other Candidate Genes.

Several other genes have been explored as candidates for MS susceptibility (Compston 1998b). Because of the associations with the HLA class II alleles reported above, logical candidates are other polymorphic genes in the same region on chromosome 6. These include the peptide transporter genes TAP1 and TAP2, the heat-shock protein, genes involved in the alternate pathway of the complement cascade, and the genes encoding tumor necrosis factors (TNF) α and β. TNF-α is particularly interesting because of its role in oligodendrocyte injury. Some studies (Sandberg-Wollheim et al. 1995; Epplen et al. 1997; Kirk et al. 1997) but not others (Fugger et al. 1990a; Roth et al. 1994; He et al. 1995; Wingerchuk et al. 1997) support an association between TNF-a alleles and risk of MS. Also on chromosome 6 is the gene for myelin oligodendrocyte glycoprotein (MOG), but no consistent associations have been reported with alleles in this region (Hilton et al. 1995; Roth et al. 1995; Rodriguez et al. 1997). Other cytokines and their receptors not encoded in the HLA region have also been investigated as potential candidates, but findings have again been inconsistent (Compston 1998b). Among the genes that could interact with the HLA class II alleles in increasing the risk of MS, the most appealing would include the alleles encoding the T-cell receptors that recognize antigens presented by the class II molecules. A strong interaction between an α-chain T-cell receptor allele (encoded on chromosome 14) and DR2 was reported in one study (Sherritt et al. 1992) but was not confirmed by others (Hashimoto et al. 1992; Hillert et al. 1992; Eoli et al. 1994). The role of polymorphisms of the β-chain remains controversial (Oksenberg et al. 1988; Beall et al. 1989; Fugger et al. 1990b; Hillert et al. 1991; Martinez-Naves et al. 1993). Another locus that could be important is the immunoglobulin heavy-chain gene on chromosome 14q. Most evidence does not appear to support a role of mitochondrial genes or genes encoding structural myelin proteins (Compston 1998b).

Genome Screening

Complete genomic screens are an alternative to the candidate gene approach. The (p.204) evidence of a strong genetic component in MS contrasts with the relatively weak associations identified thus far, suggesting that there are unknown genes with stronger effects on MS susceptibility. Genome screens are conducted among families with more than one member affected by MS (multiplex family) to identify both the regions that are more likely to harbor disease susceptibility genes and those that are very unlikely to harbor disease susceptibility genes. Four large, full genome screens have been conducted among Caucasians in the United States, Canada, the United Kingdom, and Scandinavia (Ebers et al. 1996; Haines et al. 1996; Sawcer et al. 1996; Akesson et al. 2002); smaller genome-wide screens have been conducted in Finland, Sardinia, and Italy (Kuokkanen et al. 1997; Coraddu et al. 2001; Broadley et al. 2001). No locus with overwhelming evidence of linkage emerged in any of the studies. A meta-analysis of the pooled data from three of these studies (Ebers et al. 1996; Haines et al. 1996; Sawcer et al. 1996) confirmed that no region could be clearly identified as the prime candidate for harboring a strong susceptibility gene (The American Group 2001). The strongest statistical association was on chromosome 17qll; this region encodes many potential candidates, including the gene for neurofibromatosis I (NF1), a gene that had been previously proposed as related to MS (Ferner et al. 1995). A different region of chromosome 17 (q22-q24) previously identified in the Finnish screening has been recently fine mapped using a dense set of 31 markers. This study, carried out in 22 Finnish multiplex MS families, restricted the original linkage to a region of approximately 4 centimorgan (cM) flanked by the markers D17S1792 and ATA43A10; linkage in this region was identified in 17 of the families (77.3%) (Saarela et al. 2002). The multipoint linkage analyses provided further evidence for the same 4 cM region, with a maximal multipoint NPL score of 5.98 (p < 0.0002). This region includes several functional candidate genes.

Overall, the results of genome screens support the view that heritable MS susceptibility is determined by multiple interacting genes. Stratification of the analyses by HLA-DR haplotype, refinement of the genotyping using more dense microsatellite markers as described above, and screening of regions of interest in association studies may soon lead to important discoveries.

Infection and Multiple Sclerosis

An association between infection and MS was proposed in 1883 by Pierre Marie, and an infectious cause is still being pursued, though no specific organism has been conclusively linked to MS. Two distinct theories underlie epidemiological research on infection and MS. These theories both posit that a widespread microbe, rather than a rare pathogen, causes the disease, but they differ in other aspects, as outlined below.

The first theory, based on the investigation of the Faroe Islands epidemic (Kurtzke 1993), postulates that MS is caused by a pathogen that is more common in regions of high MS incidence. According to this theory, there is a widespread transmissible agent that causes an asymptomatic persistent infection or “primary MS affection (PMSA)”; rarely, and years after the primary infection, this agent would cause neurological symptoms (MS). The second theory is the polio hypothesis (Poskanzer et al. 1963; Brody 1972), which posits that the microbe causing MS is ubiquitous and transmitted early in life in populations with low MS incidence and that being infected with this microbe after childhood increases risk of MS.

Both theories provide a reasonable explanation of the geographical distribution of MS, and could account for the reduction in risk in individuals who migrate in early childhood from high- to low-risk areas (reduced risk of exposure according to the PMSA hypothesis, infection before adolescence according to the polio hypothesis). The two theories would predict different outcomes for migration from low- to high-risk (p.205) areas—the PMSA hypothesis would predict an increase in risk, whereas the polio hypothesis would predict no increase unless the migration occurred before a child was infected—but the data on migration from low- to high-risk areas are too inconsistent to distinguish between the two theories.

If the PMSA hypothesis is true and there is a specific infectious agent that is more common in areas of high MS frequency, then some degree of space–time clustering of cases would be expected. Although several investigators have failed to find space–time clustering during childhood and adolescence among individuals in the same community who developed MS (Neutel et al. 1977; Isager et al. 1980; Larsen et al. 1985a; Riise et al. 1991; Riise 1997), the PMSA agent might be so ubiquitous that almost everyone is infected or it might be difficult to transmit among siblings. Another alternative is that clustering at the time of disease onset might be masked because of the restricted age of susceptibility, the hypothetically long asymptomatic period (at least 6 years according to Kurtzke and Hyllested 1987), and the fact that individuals with MS may no longer be infectious.

Although it has been criticized (Nathanson and Miller 1978), the polio theory retains substantial appeal (Detels et al. 1972; Alvord et al. 1987; Riise et al. 1991). Among the supportive findings are the positive association of MS with socioeconomic status (Beebe et al. 1967; Russell 1971; Kurtzke and Page 1997), the later age at infection with childhood viruses in MS cases than in controls (Granieri and Casetta 1997), and the increased risk of MS among individuals with a history of infectious mononucleosis (Operskalski et al. 1989; Lindberg et al. 1991; Martyn et al. 1993; Haahr et al. 1997; Marrie et al. 2000).

An alternative but related hypothesis is that MS is an autoimmune reaction triggered in susceptible individuals in response to infection by multiple pathogens; exposure to several infectious agents early in life would be protective, as in the polio model, but there would not be a specific agent responsible.

Whether a single specific agent or multiple microbes are involved, the prevailing view is that MS is a reaction to common pathogens that could initiate autoimmunity with or without residing in the central nervous system (CNS) (Hafler 1999; Hunter and Hafler 2000).

Candidate viruses that have been proposed as causative agents in MS include measles, rubella, mumps, herpes viruses I and II, herpes zoster, Epstein-Barr virus (EBV), cytomegalovirus, human herpes virus 6 (HHV-6), canine distemper virus, and many others (Cook et al. 1995). Most of these hypotheses originated from the finding of elevated levels of antibodies in sera from MS patients as compared with that in controls. A major limitation of this approach is that differences between cases and controls may represent a consequence of an abnormal immune response in MS patients and be of little etiological relevance (Granieri and Casetta 1997). Our view is that the data provide strong evidence, though not conclusive proof, of a causal1association between EBV and MS.

Epstein-Barr Virus

A causal association between EBV and MS was originally suggested 20 years ago (Warner and Carp 1981), primarily because of the similarities in the epidemiology of infectious mononucleosis (IM) and MS. Both diseases affect mostly young adults, follow a latitude gradient (Brodsky and Heath 1972; Hallee et al. 1974; Kurtzke 1995), are rare among populations where children are infected with EBV at an early age (Kurtzke 1995; Niederman and Evans 1997), occur at an earlier age in women (p.206) than in men (Martyn 1991; Niederman and Evans 1997), are more frequent in individuals with high socioeconomic status (Beebe et al. 1967; Russell 1971; Nye 1973; Carvalho et al. 1981; Hesse et al. 1983; Kurtzke and Page 1997), are less frequent in blacks (Brodsky and Heath 1972; Heath et al. 1972; Kurtzke et al. 1979) and Asians (Chang et al. 1979) than in whites, and occur rarely among Eskimos (Chan 1977; Melbye et al. 1984). Further, the risk of MS is significantly increased among individuals with a history of IM (Operskalski et al. 1989, Lindberg et al. 1991; Martyn et al. 1993; Haahr et al. 1997; Marrie et al. 2000; Hernán et al. 2001b).

Epstein-Barr virus is a herpesvirus that establishes a persistent latent infection in B lymphocytes (Rickinson and Kieff 1996). Antibodies against the EBV antigens persist at stable levels throughout life, and it is thus possible to determine serologically whether an individual is or is not infected with EBV. Titers of these antibodies have been found to be elevated in MS patients (Larsen et al. 1985b; Shirodaria et al. 1987) but could reflect immune disregulation in MS. Despite a high prevalence of EBV infection in most populations (typically over 90% in adults), a recent meta-analysis showed that the risk of MS is more than 10 times higher among EBV-positive than among EBV-negative subjects (Table 8–5) (Ascherio and Munch 2000). Since primary EBV infection is rare among patients with MS (Munch et al. 1998; Wandinger et al. 2000), this finding suggests that EBV itself increases the risk of MS. An important limitation of the data presented in Table 8–5 is the potential bias from control selection, as controls were not randomly sampled from the population that generated the cases. Nevertheless, the consistency of the association across studies conducted in different countries and with different control groups makes bias an unlikely explanation. Moreover, these results have now been confirmed by two more studies (Wandinger et al. 2000; Ascherio et al. 2001a), one of which (Ascherio et al. 2001a) was nested within a cohort and thus not prone to selection bias. Further evidence supporting a causal role of EBV in MS is provided by the recent report that active viral replication occurs more commonly in MS patients during exacerbations than in patients with stable disease (Wandinger et al. 2000) and by the finding in two separate studies that elevations of anti-EBV antibody titers among cases of MS precede the onset of the disease by several years (Ascherio et al. 2001a; Levin et al. 2003). In the largest of these investigations (Levin et al. 2003), the risk of developing MS was 30-fold higher among individuals in the highest category of serum IgG antibodies against the EBV nuclear antigen (EBNA) as compared with those in the lowest.

The failure to demonstrate EBV in MS plaques by in situ hybridization (Hilton et al. 1994) or polymerase chain reaction (PCR) (Morré et al. 2001) suggests that direct CNS infection is not involved. Rather, EBV could trigger an autoimmune reaction (Wucherpfennig and Strominger 1995; Vaughan et al. 1996), since infection with EBV elicits a strong, persistent, cytotoxic response. Autoimmunity could occur if some of these T cells recognized myelin epitopes.

Other Viruses

Comparisons of antibody titers for most other viruses between MS cases and controls have produced inconsistent results. Null results have been published for herpes viruses I and II (Cremer et al. 1980; Myhr et al. 1998), herpes zoster (Cremer et al. 1980; Myhr et al. 1998), cytomegalovirus (Cremer et al. 1980; Myhr et al. 1998), rubella (Cremer et al. 1980), measles (Poskanzer et al. 1980), and mumps (Cremer et al. 1980), Most importantly, a causal role of many of these viruses would not by itself explain the epidemiology of MS. For example, the lack of a decline in MS incidence following massive immunization against measles (Bansil et al. 1990) makes a primary etiological role of this virus unlikely. Human herpes virus 6, a β-herpesvirus (p.207)

Table 8–5 Epstein-Barr Seropositivity among Patients with Multiple Sclerosis Compared with Controls by Study





95% CI

Sumaya et al. 1980

157 cases with clinically definite MS

81 subjects including spouses, other household members, and laboratory personnel



Bray et al. 1983

313 cases with clinical diagnosis of MS and oliglonal bands in CSF

406 normal blood donors and patients with other nondemyelinating neurological diseases



Larsen et al. 1985b

100 cases with definite MS

100 healthy hospital staff or blood donors


Sumaya et al. 1985

104 cases with clinically definite MS

104 healthy subjects, unrelated to the cases


Shirodaria et al. 1987

26 cases with clinically definite MS

26 healthy blood donors


Ferrante et al. 1987

30 cases with definite MS

51 subjects with other diseases



Munch et al. 1997

138 caes

138 normal controls



Myhr et al. 1998

144 cases (130 definite MS; 10 probable MS; 4 possible MS)

170 hospital subjects admitted for trauma or minor surgery


Ascherio and Munch 2000

1005 cases from the 8 case–control studies listed above

1060 control from the 8 case–control studies listed above



CI, confidence intervals; CSF, cerebrospinal fluid; OR, odds ratio.

(*) Mantel-Haenszel odds ratio.

Source: Reprinted with permission from Ascherio A, Munch M. Epstein-Ban virus and multiple sclerosis. Epidemiology 2000;11:1–5.

(p.208) that exists in two variants, A and B, has been associated with MS (Wilborn et al. 1994, Soldan et al. 1997); however, other investigators could not confirm these findings (Martin et al. 1997; Fillet et al. 1998; Mayne et al. 1998; Goldberg et al. 1999; Taus et al. 2000). Recently, an association has been reported between variant A of the human herpesvirus (HHV-6) and MS (Akhyani et al. 2000; Knox et al. 2000; Soldan et al. 2000). Unlike the HHV-6 variant B, variant A may infect EBV-positive B-cell lines and activate the latent EBV genome (Cuomo et al. 1995; Flamand and Menezes 1996). These observations suggest that interactions between herpesviruses may contribute to the pathogenesis of MS.

Chlamydia pneumoniae.

An association between chlamydial infections and risk of MS was first suggested in French MS literature in the 1960s (Le Gac 1960; Jadin 1962), but the hypothesis remained relatively obscure until the recent report of a patient with rapidly progressive MS and positive cerebrospinal fluid (CSF) culture for C. pneumoniae who improved markedly after antibiotic treatment (Sriram et al. 1998). This observation was followed by a systematic study of 37 patients with MS and 27 controls with other neurological diseases (Sriram et al. 1999). C. pneumoniae was cultured from the CSF in 64% of cases and 11% of controls; moreover, 97% of the cases were positive for the presence of C. pneumoniae DNA by PCR, compared to 18% of the controls. These findings renewed the interest in Chlamydia as a possible cause of MS and several groups of investigators have attempted to reproduce them, though most have been unsuccessful (Boman et al. 2000; Hammerschlag et al. 2000; Li et al. 2000; Morré et al. 2000; Poland and Rice 2000; Pucci et al. 2000; Gieffers et al. 2001). Results of serological analyses are also conflicting (Boman et al. 2000; Krametter et al. 2001; Munger et al. 2003). The lack of consensus among C. pneumoniae–MS association studies may be due in part to the lack of standardized laboratory procedures for detection and isolation of C. pneumoniae (Peeling et al. 2000; Tompkins et al. 2000), small sample sizes, and use of different control groups.

Lifestyle and Other Factors

Cigarette Smoking

Several reports have described an aggravation of MS symptoms after smoking (Franklin and Brickner 1947; Spillane 1955; Anonymous 1964; Perkin et al. 1975; Perkin and Rose 1976; Emre and de Decker 1987, 1992), and a positive association between cigarette smoking before age of onset and risk of MS was found in a case–control study in Israel (Antonovsky et al. 1965) and in one in Canada (Ghadirian et al. 2001), although not in others (Warren et al. 1982). Most notably, a positive association was found in each of the three cohort studies that have addressed this question. Two studies in the United Kingdom reported relative risks (RR) of 1.8 (95% CI 0.8–3.6) [(Villard-Mackintosh and Vessey 1993) and 1.4 (95% CI 0.9–2.2) (Thorogood and Hannaford 1998)] among women smoking 15 or more cigarettes per day as compared with never-smokers. In the third study, conducted among U.S. nurses, the RR was 1.7 (95% CI 1.2–2.4; p < 0.01) for women who smoked 25 or more pack-years as compared with never-smokers (Hernán et al. 2001a); this association was not explained by latitude or ancestry. The combined evidence from these studies and the lack of alternative explanations suggest that smoking could increase the risk of MS. This potential effect could be due to the neurotoxic (Smith et al. 1963) or immunomodulatory (Francus et al. 1988; Sopori and Kozak 1998) effects of components of cigarette smoke. These mechanisms are indirectly supported by the association of cigarette smoking with optic neuropathy (Cuba Neuropathy Field Investigation Team 1995) and autoimmune diseases such as systemic lupus erythematosus (Hardy et al. 1998) and rheumatoid arthritis (Vessey (p.209) et al. 1987; Hernandez Avila et al. 1990; Heliovaara et al. 1993; Voigt et al. 1994; Silman et al. 1996). Finally, cigarette smoke increases the frequency and duration of respiratory infections (Graham 1990) that may contribute to the etiology of MS. If these findings are confirmed, cigarette smoking may emerge as the first modifiable risk factor for MS. Further, differences in smoking habits across populations could explain some of the variation in MS incidence, and particularly variation in the sex ratio, because of the wide range of smoking prevalence in men and women in different countries and periods.


Example 8–5 The observation in Norway of a lower prevalence of MS in coastal fishing communities, where the diet is richer in polyunsaturated fish oils, than in agricultural inland communities, where meat and dairy provide most of the fat, suggests that animal fat or saturated fat may increase the risk of MS (Swank et al. 1952). This hypothesis was supported by a significant positive correlation between consumption of calories from animal sources and prevalence of MS in a comparative study of 22 countries (Alter et al. 1974). The pooled results of three randomized trials suggested that high amounts of polyunsaturated fat may modestly reduce the severity and duration of relapses (Dworkin et al. 1984), but a 0.5 gram n-3 polyunsaturated fat supplement was not shown to be beneficial after 2 years (Bates et al. 1989). Consumption of meat (Lauer 1989a, 1989b, 1991) or milk (Agranoff and Goldberg 1974) has also been related to risk of MS in ecological investigations but not in case–control studies (Antonovsky et al. 1965; Cendrowski et al. 1969; Poskanzer et al. 1980; Butcher 1986). No significant associations were found between intake of total fat, cholesterol, specific fatty acids, dairy products, seafood, poultry, or red meats and risk of MS in a large prospective study (Zhang et al. 2000). An increase in 1% of energy from linolenic acid was associated with a lower risk (RR = 0.3, 95% CI 0.1–1.1), but this association was of borderline significance and needs confirmation (Zhang et al. 2000).

It has also been suggested that low concentrations of antioxidant enzymes leave the white matter vulnerable to oxygen free radicals and lipid peroxidation, which could be involved in the etiology of MS (Mickel 1975; Hunter et al. 1985; Clausen et al. 1988; Langemann et al. 1992). Epidemiological data on intake of antioxidants and MS are sparse. In a case–control study, intake of vitamin C was associated with a lower risk of MS, but no associations were found with vitamin E or carotene (Ghadirian et al. 1998). Intake of fruits and vegetables, which are rich in carotenoids and other dietary antioxidants, was not associated with risk of MS in several case–control studies (Antonovsky et al. 1965; Warren et al. 1982; Berr et al. 1989; Gusev et al. 1996; Ghadirian et al. 1998). In the only prospective investigation, no associations were found between intake of vitamin E, vitamin C, or specific carotenoids and risk of MS (Zhang et al. 2001).

A protective role of vitamin D has also been proposed, and could explain in part the latitude gradient, but support for this hypothesis is so far limited to experimental animal data (Hayes 2000). Because of the consistent contribution of dietary intake and sunlight to vitamin D status, the potential protective effect of this vitamin will be better assessed in prospective serological studies. Diet could also be important because of specific antigenic stimulation rather than the effect of specific nutrients. In an experiment in rodents, the immune response to a cow milk protein, butyrophilin, led to encephalitis through antigenic mimicry with MOG (Stefferl et al. 2000; Winer et al. 2001). These studies suggest a mechanism by which consumption of milk products could affect the risk of MS, but, as discussed above, epidemiological evidence supporting this association is weak.

In summary, a role of diet in the etiology of MS remains unproven but cannot be (p.210) excluded, as most reported investigations were retrospective and prone to several sources of bias (Willett 1998). A potential beneficial effect of linolenic acid and vitamin D should be further addressed in larger prospective investigations.

Hormonal Factors

Sex hormones modulate the immune response and could thus influence the onset and progression of MS (Grossman 1985; Whitacre et al. 1999, 2001). Low levels of estrogens seem to favor a pro-inflammatory type 1 response in T cells, whereas high levels of estrogens and progesterone favor a type 2 response. In animals, exogenous estrogens suppress experimental autoimmune encephalomyelitis (Kim et al. 1999; Bebo et al. 2001). Relapses of MS are rare during pregnancy, when levels of circulating estrogens are high, but increase in frequency during the puerperium, so that overall there seems to be little impact of pregnancy on the progression of the disease (Confavreux et al. 1998). Also, there is no evidence in prospective studies that parity or age at first birth affects the risk of MS (Vessey et al. 1976; Thorogood and Hannaford 1998; Hernán et al. 2000). The effect of use of oral contraceptives on risk of MS was also investigated in the same cohorts, but no significant associations were found (Vessey et al. 1976; Thorogood and Hannaford 1998; Hernán et al. 2000). The largest of these investigations comprised over 200,000 women followed for up to 18 years and among whom 315 cases of definite or probable MS were documented. The RR of MS among women who used oral contraceptives for 8 years or more as compared with never-users was 1.2 (95% CI 0.8–1.2) (Hernán et al. 2000). Therefore, a protective effect of oral contraceptives seems unlikely.

Other Factors

It has been proposed that trauma could increase the risk of MS by disrupting the blood–brain barrier (Poser 1987), but overall little evidence supports this hypothesis (Compston 1998a). Recently, concerns were raised about the possibility that administration of the hepatitis B vaccine may increase the risk of MS (Marshall 1998; Touze et al. 2000), but no associations between the vaccine and risk of MS (Ascherio et al. 2001a) or MS relapses (Confavreux et al. 2001) were found in prospective studies. Other environmental factors investigated in relation to MS include exposures to heavy metals, solvents, and other contaminants (Eastman et al. 1973; Stein et al. 1987; Irvine et al. 1988; Ingalls 1989; Landtblom 1997); their roles remain uncertain.

Prognostic Studies

Since MS is a chronic and often progressive disease with a variable course, the investigation of factors that may modify the course of the illness and its effects on the quality of life is of great importance. Randomized trials typically include selected groups of patients, and the follow-up rarely extends beyond a few years. Thus, complementary information from large, long-term, prospective studies is important.

Prognostic Factors

Factors predictive of shorter survival after the onset of MS include late age at onset (Riise et al. 1988; Poser et al. 1989; Midgard et al. 1995; Wallin et al. 2000), male gender (McAlpine 1961; Poser et al. 1986; Wynn et al. 1990; Wallin et al. 2000), an initial progressive course (Poser et al. 1986; Phadke 1987; Riise et al. 1988; Midgard et al. 1995), and absence of sensory symptoms at onset (Midgard et al. 1995). The shorter life expectancy observed among males and those with an older age at MS onset may in part reflect general population trends. Among U.S. Army veterans, there was a significant difference in survival rates between men and women with MS, but this difference was similar to that found among U.S. Army personnel without MS (Wallin et al. 2000). In Olmsted County, Minnesota, however, MS was associated (p.211) with higher mortality in men, but not in women (Wynn et al. 1990).

There is some disagreement as to whether early age at onset and female gender are prognostic for a favorable disability outcome independent of initial disease course (McAlpine 1961; Leibowitz and Alter 1970; Kurtzke et al. 1977; Confavreux et al. 1980; Clark et al. 1982; Detels et al. 1982; Poser et al. 1982; Verjans et al. 1983; Visscher et al. 1984; Thompson et al. 1986; Lauer and Firnhaber 1987; Phadke 1990; Weinshenker et al. 1991; Riise et al. 1992; Runmarker and Andersen 1993; Amato et al. 1999; Hammond et al. 2000b; Liguori et al. 2000). In some studies, gender was not a significant predictor of disability outcome after adjustment for type of MS, onset symptoms, and disease duration, whereas age at onset remained significant (Weinshenker et al. 1991a; Hammond et al. 2000b). In others, neither gender nor age at onset was a significant predictor of disability after controlling for type of MS and symptoms (Amato et al. 1999). Onset with motor symptoms has been associated with a poor MS course as compared with onset with sensory symptoms (Clark et al. 1982; Visscher et al. 1984; Sanders et al. 1986; Phadke 1990; Amato et al. 1999) or optic neuritis (McAlpine 1961; Poser et al. 1982; Sanders et al. 1986; Optic Neuritis Study Group 1997). These differences remained significant after adjustment for gender, age at onset, and disease course (Weinshenker et al. 1991; Amato et al. 1999; Hammond et al. 2000b). Some studies, however, have found no predictive value of onset symptoms (Kurtzke et al. 1977; Confavreux et al. 1980; Lauer and Firnhaber 1987) and one study reported a favorable prognosis with motor symptoms at onset (Leibowitz and Alter 1970). The HLA-DR2 genotype has also been associated with a more severe disease course in some studies (Jersild et al. 1973; Engell et al. 1982; Duquette et al. 1985) but not in others (Poser et al. 1981; Madigand et al. 1982; Dejaegher et al. 1983). A recent prospective study in The Netherlands assessed whether systemic infections contributed to the natural course of exacerbations (Buljevac et al. 2002). The investigators observed 167 infections and 145 exacerbations in 73 patients with relapsing-remitting MS during 6466 patient weeks. Exacerbations increased twofold during a predefined period of 2 weeks before until 5 weeks after onset of a clinical infection (mainly upper airway infections). Systemic infections may lead to more sustained damage than other exacerbations. Other factors found predictive of a favorable disability outcome include monoregional onset symptoms (Runmarker and Andersen 1993), complete recovery from relapses (Runmarker and Andersen 1993), and longer interval between early relapses (Confavreux et al. 1980; Thompson et al. 1986; Phadke 1990; Amato et al. 1999; Myhr et al. 2001).

Future Directions and Conclusions

Epidemiological studies support the existence of a strong but still unidentified environmental determinant(s) of MS. Less certain is whether this is an infectious or a noninfectious agent. Several aspects in the epidemiology of MS are consistent with the hypothesis that MS is an autoimmune reaction to infection with one or more microbes, and there is strong, albeit nonconclusive, evidence that EBV alone or in combination with other viruses is involved in the etiology of MS. This hypothesis does not exclude a role of noninfectious factors, such as cigarette smoking and diet.

Familial studies indicate that the aggregation of MS within families is clearly due to genetic and not to environmental factors. The risk of MS among genetically predisposed individuals appears to be over 100-fold higher than in the general population, yet the responsible genes have not been identified. It is now recognized that two or more genes interact in increasing the risk of MS; only large investigations and pooling of data from different studies will provide the power needed to detect these epistatic effects. Advances in both genome-screening (p.212) techniques and statistical methods will be critical to this search, but its success will also depend on the complexity of nature, as genetic heterogeneity and poligenicity could present formidable obstacles. The research on environmental and genetic determinants will probably converge and spur each other, as interactions are most likely to exist between genes and environment.


Bibliography references:

Agranoff BW, Goldberg D. Diet and the geographical distribution of multiple sclerosis. Lancet 1974;2:1061–1066.

Akesson E, Oturai A, Berg J, et al. A genomewide screen for linkage in Nordic sib-pairs with multiple sclerosis. Genes Immun 2002;3:279–285.

Akhyani N, Berti R, Brennan MB, et al. Tissue distribution and variant characterization of human herpesvirus (HHV)-6: increased prevalence of HHV-6A in patients with multiple sclerosis. J Infect Dis 2000;182:1321–1325.

Alter M, Halpern L, Kurland LT, et al. Multiple sclerosis in Israel. Arch Neurol 1962;7:253–263.

Alter M, Kahana E, Loewenson R. Migration and risk of multiple sclerosis. Neurology 1978;28:1089–1093.

Alter M, Leibowitz U, Speer J. Risk of multiple sclerosis related to age at immigration to Israel. Arch Neurol 1966;15:234–237.

Alter M, Yamoor M, Harshe M. Multiple sclerosis and nutrition. Arch Neurol 1974;31:267–272.

Alvord EC Jr, Jahnke U, Fischer EH, et al. The multiple causes of multiple sclerosis: the importance of age of infections in childhood. J Child Neurol 1987;2:313–321.

Amato MP, Ponziani G, Bartolozzi ML, Siracusa G. A prospective study on the natural history of multiple sclerosis: clues to the conduct and interpretation of clinical trials. J Neurol Sci 1999;168:96–106.

American Group. A meta-anlaysis of genomic screens in multiple sclerosis. Mult Scler 2001;7:3–11.

Anonymous. Smoking and multiple sclerosis. BMJ 1964;1:773.

Antonovsky A, Leibowitz U, Smith HA, et al. Epidemiologic study of multiple sclerosis in Israel. Arch Neurol 1965;13:183–193.

Ascherio A, Munger KL, Lennette ET, et al. Epstein-Barr virus antibodies and risk of multiple sclerosis: a prospective study. JAMA 2001a;286:3083–3088.

Ascherio A, Munch M. Epstein-Barr virus and multiple sclerosis. Epidemiology 2000;11:1–5.

Ascherio A, Zhang SM, Hernán MA, et al. Hepatitis B vaccination and the risk of multiple sclerosis. N Engl J Med 2001b;344:327–332.

Bansil S, Troiano R, Dowling PC, Cook SD. Measles vaccination does not prevent multiple sclerosis. Neuroepidemiology 1990;9:248–254.

Bates D, Cartlidge NEF, French JM, et al. A double-blind controlled trial of long chain n-3 polyunsaturated fatty acids in the treatment of multiple sclerosis. J Neurol Neurosurg Psychiatry 1989;52:18–22.

Baum HM, Rothschild BB. The incidence and prevalence of reported multiple sclerosis. Ann Neurol 1981;10:420–428.

Beall SS, Concannon P, Charmley P, et al. The germline repertoire of T cell receptor β-chain genes in patients with chronic progressive multiple sclerosis. J Neuroimmunol 1989;21:59–66.

Bebo BF Jr, Fyfe-Johnson A, Adlard K, et al. Low-dose estrogen therapy ameliorates experimental autoimmune encephalomyelitis in two different inbred mouse strains. J Immunol 2001;166:2080–2089.

Beebe G, Kurtzke JF, Kurland LT, et al. Studies on the natural history of multiple sclerosis. 3. Epidemiologic analysis of the Army experience in World War II. Neurology 1967;17:1–17.

Berr C, Puel J, Clanet M, et al. Risk factors in multiple sclerosis: a population-based case–control study in Hautes-Pyrenees, France. Acta Neurol Scand 1989;80:46–50.

Bobowick AR, Kurtzke JF, Brody JA, et al. Twin study of multiple sclerosis: an epidemiologic inquiry. Neurology 1978;28:978–987.

Boman J, Roblin PM, Sundstrom P, et al. Failure to detect Chlamydia pneumoniae in the central nervous system of patients with MS. Neurology 2000;54:265.

Bray PF, Bloomer LC, Salmon VC, et al. Epstein-Barr virus infection and antibody synthesis in patients with multiple sclerosis. Arch Neurol 1983;40:406–408.

Broadley S, Sawcer S, D’Alfonso S, et al. A genome screen for multiple sclerosis in Italian families. Genes Immun 2001;2:205–210.

Brodsky AL, Heath CW Jr. Infectious mononucleosis: epidemiologic patterns at United States colleges and universities. Am J Epidemiol 1972;96:87–93.

Brody JA. Epidemiology of multiple sclerosis and a possible virus aetiology. Lancet 1972;2:173–176.

Buljevac D, Flach HZ, Hop WCJ, et al. Prospective study on the relationship between infections and multiple sclerosis exacerbations. Brain 2002;125:952–960.

(p.213) Bulman DE, Ebers GC. The geography of multiple sclerosis reflects genetic susceptibility. J Tropic Geograph Neurol 1992;2:66–72.

Butcher PJ. Milk consumption and multiple sclerosis—an etiological hypothesis. Med Hypothes 1986;19:169–178.

Carton H, Vlietinck R, Debruyne J, et al. Risks of multiple sclerosis in relatives of patients in Flanders, Belgium. J Neurol Neurosurg Psychiatry 1997;62:329–333.

Carvalho RP, Evans AS, Pannuti CS, et al. EBV infections in Brazil. III—Infectious mononucleosis. Rev Inst Med Trop Sao Paulo 1981;23:167–172.

Cendrowski W, Wender M, Dominik W, et al. Epidemiological study of multiple sclerosis in Western Poland. Eur Neurol 1969;2:90–108.

Chan WW. Eskimos and multiple sclerosis. Lancet 1977;1:1370.

Chang RS, Char DFB, Jones JH, Halstead SB. Incidence of infectious mononucleosis at the Universities of California and Hawaii. J Infect Dis 1979;140:479–486.

Clark VA, Detels R, Visscher BR, et al. Factors associated with a malignant or benign course of multiple sclerosis. JAMA 1982;248:8568–8560.

Clausen J, Jensen GE, Nielsen SA. Selenium in chronic neurologic diseases. Biol Trace Element Res 1988;15:179–203.

Compston A. Distribution of multiple sclerosis. In: Compston A, ed. McAlpine’s Multiple Sclerosis, 3rd edition. New York: Churchill Livingstone, 1998a, pp 63–100.

Compston A. Genetic susceptibility to multiple sclerosis. In: Compston A, ed. McAlpine’s Multiple Sclerosis, 3rd edition. New York: Churchill Livingstone, 1998b, pp 101–142.

Confavreux C, Aimard G, Devic M. Course and prognosis of multiple sclerosis assessed by the computerized data processing of 349 patients. Brain 1980;103:281–300.

Confavreux C, Hutchinson M, Hours MM, et al. Rate of pregnancy-related relapse in multiple sclerosis. Pregnancy in multiple sclerosis group. N Engl J Med 1998;339:285–291.

Confavreux C, Suissa S, Saddier P, et al. Vaccinations and the risk of relapse in multiple sclerosis. Vaccines in Multiple Sclerosis Study Group. N Engl J Med 2001;344:319–326.

Cook SD, Rohowsky-Kochan C, Bansil S, Dowling PC. Evidence for multiple sclerosis as an infectious disease. Acta Neurol Scand 1995;161(Suppl):34–42.

Coraddu F, Sawcer S, D’Alfonso S, et al. A genome screen for multiple sclerosis in Sardinian multiplex families. Eur J Hum Genet 2001;9:621–626.

Cremer NE, Johnson KP, Fein G, Likosky WH. Comprehensive viral immunology of multiple sclerosis. II. Analysis of serum and CSF antibodies by standard serologic methods. Arch Neurol 1980;37:610–615.

Cuba Neuropathy Field Investigation Team. Epidemic optic neuropathy in Cuba—clinical characterization and risk factors. The Cuba Neuropathy Field Investigation Team. N Engl J Med 1995;333:1176–1182.

Cuomo L, Angeloni A, Zompetta C, et al. Human herpesvirus 6 variant A, but not variant B, infects EBV-positive B lymphoid cells, activating the latent EBV genome through a BZLF-1-dependent mechanism. AIDS Res Hum Retroviruses 1995;11:1241–1245.

Davenport CB. Multiple sclerosis from the standpoint of geographic distribution and race. Arch Neurol Psychiatry 1922;8:51–58.

Dean G. Annual incidence, prevalence, and mortality of multiple sclerosis in white South-African-born and in white immigrants to South Africa. BMJ 1967;2:724–730.

Dean G. Was there an epidemic of multiple sclerosis in the Faroe Islands? Neuroepidemiology 1988;7:165–167.

Dean G, Brady R, McLoughlin H, et al. Motor neurone disease and multiple sclerosis among immigrants to Britain. Br J Prev Soc Med 1977;31:141–147.

Dean G, Grimaldi G, Kelly R, Karhausen L. Multiple sclerosis in southern Europe. I: Prevalence in Sicily in 1975. J Epidemiol Community Health 1979;33:107–110.

Dean G, McLoughlin H, Brady R, et al. Multiple sclerosis among immigrants in Greater London. BMJ 1976;1:861–864.

Dejaegher L, de Bruyere M, Ketelaer P, Carton H. HLA antigens and progression of multiple sclerosis. Part II. J Neurol 1983;229:167–174.

de Jong BA, Huizinga TWJ, Zanelli E et al. Evidence for additional genetic risk indicators of relapse-onset MS within the HLA region. Neurology 2002;59:549–555.

Detels R, Brody JA, Edgar AH. Multiple sclerosis among American, Japanese and Chinese immigrants to California and Washington. J Chron Dis 1972;25:3–10

Detels R, Clark VA, Valdiviezo NL, et al. Factors associated with a rapid course of multiple sclerosis. Arch Neurol 1982;39:337–341.

Detels R, Visscher BR, Haile RW, et al. Multiple sclerosis and age at migration. Am J Epidemiol 1978;108:386–393.

Duquette P, Decary F, Pleines J, et al. Clinical sub-groups of multiple sclerosis in relation to HLA: DR alleles as possible markers of disease progression. Can J Neurol Sci 1985;12:106–110.

(p.214) Dworkin RH, Bates D, Paty DW. Linoleic acid and multiple sclerosis. Neurology 1984;34:1141–1445.

Eastman R, Sheridan J, Poskanzer DC. Multiple sclerosis clustering in a small Massachusetts community, with possible common exposure 23 years before onset. N Engl J Med 1973;289:793–794.

Ebers G. Natural history of multiple sclerosis. In: Compston A, ed. McAlpine’s Multiple Sclerosis, 3rd edition. New York: Churchill Livingstone, 1998, pp 191–221.

Ebers GC, Bulman DE, Sadovnick AD, et al. A population-based study of multiple sclerosis in twins. N Engl J Med 1986;315:1638–1642.

Ebers GC, Kukay K, Bulman DE, et al. A full genome search in multiple sclerosis. Nat Genet 1996;13:472–476.

Ebers GC, Sadovnick AD, Risch NJ. A genetic basis for familial aggregation in multiple sclerosis. Canadian Collaborative Study Group. Nature 1995;377:150–151.

Ebers GC, Yee IM, Sadovnick AD, Duquette P. Conjugal multiple sclerosis: population-based prevalence and recurrence risks in offspring. Canadian Collaborative Study Group. Ann Neurol 2000;48:927–931.

Elian M, Dean G. Multiple sclerosis among the United Kingdom–born children of immigrants from the West Indies. J Neurol Neurosurg Psychiatry 1987;50:327–332.

Elian M, Dean G. Motor neuron disease and multiple sclerosis among immigrants to England from the Indian subcontinent, the Caribbean, and East and West Africa. J Neurol Neurosurg Psychiatry 1993;56:454–457.

Elian M, Nightingale S, Dean G. Multiple sclerosis among United Kingdom–born children of immigrants from the Indian subcontinent, Africa and the West Indies. J Neurol Neurosurg Psychiatry 1990;53:906–911.

Emre M, de Decker C. Nicotine and CNS. Neurology 1987;37:1887–1888.

Emre M, de Decker C. Effects of cigarette smoking on motor functions in patients with multiple sclerosis. Arch Neurol 1992;49:1243–1247.

Engell T. A clinico-pathoanatomical study of multiple sclerosis diagnosis. Acta Neurol Scand 1988;78:39–44.

Engell T, Raun NE, Thomsen M, Platz P. HLA and heterogeneity of multiple sclerosis. Neurology 1982;32:1043–1046.

Eoli M, Wood NW, Kellar-Wood HF, et al. No linkage between multiple sclerosis and the T cell receptor alpha chain locus. J Neurol Sci 1994;124:32–37.

Epplen C, Jäckel S, Santos EJM, et al. Genetic predisposition to multiple sclerosis as revealed by immunoprinting. Ann Neurol 1997;41:341–352.

Ferner RE, Hughes RA, Johnson MR. Neurofibromatosis 1 and multiple sclerosis. J Neurol Neurosurg Psychiatry 1995;58:582–585.

Ferrante P, Castellani P, Barbi M, Bergamini F. The Italian Cooperative Multiple Sclerosis case–control study: preliminary results on viral antibodies. Ital J Neurol Sci 1987;Suppl 6:45–50.

Fillet A-M, Lozeron P, Agut H, et al. HHV-6 and multiple sclerosis. Nat Med 1998;4:537.

Flamand L, Menezes J. Cyclic AMP–responsive element-dependent activation of Epstein-Barr virus zebra promoter by human herpesvirus 6. J Virol 1996;70:1784–1791.

Fog M, Kyllested K. Prevalence of disseminated sclerosis in the Faroes, the Orkneys and Shetland. Acta Neurol Scand 1966;42(Suppl 19):9–11.

Forbes RB, Swingler RJ. An epidemiologic study of multiple sclerosis in Northern Ireland. Neurology 1999;52:215–216.

Francus T, Klein RF, Staiano-Coico L, et al. Effects of tobacco glycoprotein (TGP) on the immune system. II. TGP stimulates the proliferation of human T cells and the differentiation of human B cells into Ig secreting cells. J Immunol 1988;140:1823–1829.

Franklin CR, Brickner RM. Vasospasm associated with multiple sclerosis. Arch Nerurol Psychiatry 1947;58:125–162.

Fugger L, Morling N, Sandberg-Wollheim M, et al. Tumor necrosis factor alpha gene polymorphism in multiple sclerosis and optic neuritis. J Neuroimmunol 1990a;27:85–88.

Fugger L, Sandberg-Wollheim M, Morling N, et al. The germline repertoire of T-cell receptor beta chain genes in patients with relapsing/remitting multiple sclerosis or optic neuritis. lmmunogenetics 1990b;31:278–280.

Gale CR, Martyn CN. Migrant studies in multiple sclerosis. Prog Neurobiol 1995;47:425–448.

Ghadirian P, Dadgostar B, Azani R, Maison-neuve P. A case–control study of the association between sociodemographic, lifestyle and medical history factors and multiple sclerosis. Can J Public Health 2001;92:281–285.

Ghadirian P, Jain M, Ducic S, et al. Nutritional factors in the aetiology of multiple sclerosis: a case–control study in Montreal, Canada. Int J Epidemiol 1998;27:845–852.

Gieffers J, Pohl D, Treib J, et al. Presence of Chlamydia pneumoniae DNA in the cerebral spinal fluid is a common phenomenon in a variety of neurological diseases and not restricted to multiple sclerosis. Ann Neurol 2001;49:585–589.

(p.215) Giordano M, D’Alfonso S, Momigliano-Richiardi P. Genetics of multiple sclerosis: linkage and association studies. Am J Pharmacogenomics 2002;2:37–58.

Goldberg SH, Albright AV, Lisak RP, González-Scarano F. Polymerase chain reaction analysis of human herpesvirus-6 sequences in the sera and cerebrospinal fluid of patients with multiple sclerosis. J Neurovirol 1999;5:134–139.

Graham NM. The epidemiology of acute respiratory infections in children and adults: a global perspective. Epidemiol Rev 1990;12:149–178.

Granieri E, Casetta I. Part III: selected reviews. Common childhood and adolescent infections and multiple sclerosis. Neurology 1997;49(Suppl 2):S42–S54.

Grønning M, Riise T, Kvale G, et al. Incidence of multiple sclerosis in Hordaland, western Norway: a fluctuating pattern. Neuroepidemiology 1991;10:53–61.

Grossman CJ. Interactions between the gonadal steroids and the immune system. Science 1985;227:257–261.

Gusev E, Boiko A, Lauer K, et al. Environmental risk factors in MS: a case–control study in Moscow. Acta Neurol Scand 1996;94:386–394.

Haahr S, Koch-Henriksen N, Moeller-Larsen A, et al. Increased risk of multiple sclerosis after late Epstein-Barr virus infection. Acta Neurol Scand 1997;169(Suppl):70–75.

Hafler DA. The distinction blurs between an autoimmune versus microbial hypothesis in multiple sclerosis. J Clin Invest 1999;104:527–529.

Haines JL, Ter-Minassian M, Bazyk A, et al. A complete genomic screen for multiple sclerosis underscores a role for the major histocompatability complex. The Multiple Sclerosis Genetics Group. Nat Genet 1996;13:469–471.

Hallee TJ, Evans AS, Niederman JC, et al. Infectious mononucleosis at the United States Military Academy. A prospective study of a single class over four years. Yale J Biol Med 1974;3:182–195.

Hammerschlag MR, Ke Z, Lu F, et al. Is chlamydia pneumoniae present in brain lesions of patients with multiple sclerosis? J Clin Microbiol 2000;38:4274–4276.

Hammond SR, English DR, McLeod JG. The age-range of risk of developing multiple sclerosis: evidence from a migrant population in Australia. Brain 2000a;123:968–974.

Hammond SR, McLeod JG, Macaskill P, English DR. Multiple sclerosis in Australia: prognostic factors. J Clin Neurosci 2000b;7:16–19.

Hammond SR, McLeod JG, Millingen KS, et al. The epidemiology of multiple sclerosis in three Australian cities: Perth, Newcastle and Hobart. Brain 1988;111:1–25.

Hardy CJ, Palmer BP, Muir KR, et al. Smoking history, alcohol consumption, and systemic lupus erythematosus: a case–control study. Ann Rheum Dis 1998;57:451–455.

Hashimoto LL, Mak TW, Ebers GC. T cell receptor alpha chain polymorphisms in multiple sclerosis. J Neuroimmunol 1992;40:41–48.

Hayes CE. Vitamin D: a natural inhibitor of multiple sclerosis. Proc Nutr Soc 2000;59:531–535.

He B, Navikas V, Lundahl J, et al. Tumor necrosis factor alpha-308 alleles in multiple sclerosis and optic neuritis. J Neuroimmunol 1995;63:143–147.

Heath CW Jr., Brodsky AL, Potolsky AI. Infectious mononucleosis in a general population. Am J Epidemiol 1972;95:46–52.

Heliovaara M, Aho K, Aromaa A, Knekt P, Reunanen A. Smoking and risk of rheumatoid arthritis. J Rheumatol 1993;20:1830–1835.

Heltberg A, Holm NV. Concordance in Twins and Recurrence in Sibships in Multiple Sclerosis. London: Lancet, 1982.

Hernán M, Hohol M, Olek M, et al. Oral contraceptives and the incidence of multiple sclerosis. Neurology 2000;55:848–854.

Hernán MA, Olek MJ, Ascherio A. Geographic variation of MS incidence in two prospective studies of US women. Neurology 1999;53:1711–1718.

Hernán M, Olek M, Ascherio A. Cigarette smoking and incidence of multiple sclerosis. Am J Epidemiol 2001a;154:69–74.

Hernán M, Zhang SM, Lipworth L, et al. Multiple sclerosis and age at infection with common viruses. Epidemiology 2001b;12:301–306.

Hernandez Avila M, Liang MH, Willett WC, et al. Reproductive factors, smoking, and the risk for rheumatoid arthritis. Epidemiology 1990;1:285–291.

Herndon RM. The changing pattern of misdiagnosis in multiple sclerosis. In: Hernson RM, Seil FS, eds. Multiple Sclerosis: Current Status of Research and Treatment. New York: Demos, 1994, pp 149–155.

Hesse J, Ibsen KK, Krabbe S, Uldall P. Prevalence of antibodies to Epstein-Barr Virus (EBV) in childhood and adolescence in Denmark. Scand J Infect Dis 1983;15:335–338.

Hillert J, Leng C, Olerup O. No association with germline T cell receptor beta-chain gene alleles or haplotypes in Swedish patients with multiple sclerosis. J Neuroimmunol 1991;32:141–147.

(p.216) Hillert J, Leng C, Olerup O. T-cell receptor alpha chain germline gene polymorphisms in multiple sclerosis. Neurology 1992;42:80–84.

Hilton AA, Slavin AJ, Hilton DJ, Bernard CC. Characterization of cDNA and genomic clones encoding human myelin oligodendrocyte glycoprotein. J Neurochem 1995;65:309–318.

Hilton DA, Love S, Fletcher A, Pringle JH. Absence of Epstein-Barr virus RNA in multiple sclerosis as assessed by in situ hybridisation. J Neurol Neurosurg Psychiatry 1994;57:975–976.

Hook EB, Regal RR. Capture–recapture methods in epidemiology: methods and limitations. Epidemiol Rev 1995;17:243–264.

Hunter MIS, Nlemadim BC, Davidson DLW. Lipid peroxidation products and antioxidant proteins in plasma and cerebrospinal fluid from multiple sclerosis patients. Neurochem Res 1985;10:1645–1652.

Hunter SF, Hafler DA. Ubiquitous pathogens: links between infection and autoimmunity in MS? Neurology 2000;55:164–165.

Ingalls TH. Clustering of multiple sclerosis in Galion, Ohio, 1982–1985. Am J Forensic Med Pathol 1989;10:213–215.

Irvine DG, Schiefer HB, Hader WJ. Geotoxicology of multiple sclerosis: the Henribourg, Saskatchewan, cluster focus. II. The soil. Sci Total Environ 1988;77:175–188.

Isager H, Larsen S, Hyllested K. School contact between patients with multiple sclerosis. Int J Epidemiol 1980;9:145–147.

Jacobs L, Salazar AM, Herndon RM, et al. Multicentre double-blind study of effect of intrathecially administered natural human fibroblast interferon on exacerbations of multiple sclerosis. Lancet 1986;2:1411–1413.

Jadin J. Rickettsial diseases and multiple sclerosis. Ann Soc Belge Med Trop 1962;42:321–345.

Jersild C, Fog T, Hansen GS, et al. Histocompatibility determinants in multiple sclerosis, with special reference to clinical course. Lancet 1973;2:1221–1225.

Kahana E, Zilber N, Abramson JH, et al. Multiple sclerosis: genetic versus environmental aetiology: epidemiology in Israel updated. Neurology 1994;241:341–346.

Kalman B, Lublin FD. The genetics of multiple sclerosis. A review. Biomed Pharmacother 1999;53:358–370.

Kim S, Liva SM, Dalai MA, et al. Estriol ameliorates autoimmune demyelinating disease: implications for multiple sclerosis. Neurology 1999;52:1230–1238.

Kinnunen E, Juntunen J, Ketonen L, et al. Genetic susceptibility to multiple sclerosis. A co-twin study of a nationwide series. Arch Neurol 1988;45:1108–1111.

Kirk CW, Droogan AG, Hawkins SA, et al. Tumour necrosis factor microsatellites show association with multiple sclerosis. J Neurol Sci 1997;47:21–25.

Knox KK, Brewer JH, Henry JM, et al. Human herpesvirus 6 and multiple sclerosis: systemic active infections in patients with early disease. Clin Infect Dis 2000;31:894–903.

Koch-Henriksen N. Multiple sclerosis in Scandinavia and Finland. Acta Neurol Scand 1995;161:55–59.

Koch-Henriksen N, Hyllested K. Epidemiology of multiple sclerosis: incidence and prevalence rates in Denmark 1948–64 based on the Danish Multiple Sclerosis Registry. Acta Neurol Scand 1988;78:369–380.

Krametter D, Niederwieser G, Berghold A, et al. Chlamydia pneumoniae in multiple sclerosis: humoral immune responses in serum and cerebrospinal fluid and correlation with disease activity marker. Mult Scler 2001;7:13–18.

Kuokkanen S, Gschwend M, Rioux JD, et al. Genomewide scan of multiple sclerosis in Finnish multiplex families. Am J Hum Genet 1997;61:1379–1387.

Kurland LT. The frequency and geographic distribution of multiple sclerosis as indicated by mortality statistics and morbidity surveys in the United States and Canada. Am J Hyg 1952;55:457–476.

Kurland LT, Stasio A, Reed D. An appraisal of population studies of multiple sclerosis. Ann NY Acad Sci 1965;122:520–541.

Kurtzke JF. Multiple sclerosis among immigrants. BMJ 1976;1:1527–1528.

Kurtzke JF. Multiple sclerosis in the Faroe Islands. II. Clinical update, transmission, and the nature of MS. Neurology 1986;36:307–328.

Kurtzke JF. Multiple sclerosis: what’s in a name? Neurology 1988;38:309–316.

Kurtzke JF. Epidemiologic evidence for multiple sclerosis as an infection. Clin Microbiol Rev 1993;6:382–427.

Kurtzke JF. MS Epidemiology worldwide. One view of current status. Acta Neurol Scand 1995;161(Suppl):23–33.

Kurtzke JF, Beebe GW, Nagler B, et al. Studies on the natural history of multiple sclerosis—8. Early prognostic features of the later course of the illness. J Chron Dis 1977;30:819–830.

Kurtzke JF, Beebe GW, Norman JE. Epidemiology of multiple sclerosis in U.S. veterans: 1. Race, sex, and geographic distribution. Neurology 1979;29:1228–1235.

(p.217) Kurtzke JF, Beebe GW, Norman JE. Epidemiology of multiple sclerosis in US veterans: III. Migration and the risk of MS. Neurology 1985;35:672–678.

Kurtzke JF, Dean G, Botha DPJ. A method for estimating the age at immigration of white immigrants to South Africa, with an example of its importance. S Afr Med J 1970;44:663–669.

Kurtzke JF, Delasnerie-Laupretre N, Wallin MT. Multiple sclerosis in North African migrants to France. Acta Neurol Scand 1998;98:302–309.

Kurtzke JF, Heltberg A. Multiple sclerosis in the Faroe Islands: an epitome. J Clin Epidemiol 2001;54:1–22.

Kurtzke JF, Hyllested K. Multiple sclerosis in the Faroe Islands. I. Clinical and epidemiological features. Ann Neurol 1979;5:6–21.

Kurtzke JF, Hyllested K. Multiple sclerosis in the Faroe Islands. III. An alternative assessment of the three epidemics. Acta Neurol Scand 1987;76:317–339.

Kurtzke JF, Page WF. Epidemiology of multiple sclerosis in US veterans: VII. Risk factors for MS. Neurology 1997;48:204–213.

Landtblom AM. Exposure to organic solvents and multiple sclerosis. Neurology 1997;49:S70–S74.

Langemann H, Kabiersch A, Newcombe J. Measurement of low-molecular-weight antioxidants, uric acid, tyrosine and tryptophan in plaques and white matter from patients with multiple sclerosis. Eur Neurol 1992;32:248–252.

Larsen JP, Riise T, Nyland H, et al. Clustering of multiple sclerosis in the county of Hordaland, Western Norway. Acta Neurol Scand 1985a;71:390–395.

Larsen PD, Bloomer LC, Bray PF. Epstein-Barr nuclear antigen and viral capsid antigen antibody titers in multiple sclerosis. Neurology 1985b;35:435–438.

Lassmann H, Bruck W, Lucchinetti C. Heterogeneity of multiple sclerosis pathogenesis: implications for diagnosis and therapy. Trends Mol Med 2001;7:115–121.

Lauer K. Dietary changes in temporal relation to multiple sclerosis in the Faroe Islands: an evaluation of literary sources. Neuroepidemiology 1989a;8:200–206.

Lauer K. Multiple sclerosis in relation to meat preservation in France and Switzerland. Neuroepidemiology 1989b;8:308–315.

Lauer K. The food pattern in geographical relation to the risk of multiple sclerosis in the Mediterranean and Near East region. J Epidemiol Community Health 1991;45:251–252.

Lauer K, Firnhaber W. Epidemiological investigations into multiple sclerosis in Southern Hesse. V. Course and prognosis. Acta Neurol Scand 1987;76:12–17.

Le Gac P. The treatment of multiple sclerosis caused by rickettsia and neorickettsia. J Med Bordeaux 1960;137:577–589.

Leibowitz U, Alter M. Clinical factors associated with increased disability in multiple sclerosis. Acta Neurol Scand 1970;46:53–70.

Li W, Cook S, Blumberg B, Dowling P. Chlamydia pneumoniae sequence frequently present in both MS and control spinal fluid. Neurology 2000:A165.

Liguori M, Marrosu MG, Pugliatti M, et al. Age at onset in multiple sclerosis. Neurol Sci 2000;21:S825–S829.

Lindberg C, Andersen O, Vahlne A, et al. Epidemiological investigation of the association between infectious mononucleosis and multiple sclerosis. Neuroepidemiology 1991;10:62–65.

Lublin FD, Reingold SC. Defining the clinical course of multiple sclerosis: results of an international survey. National Multiple Sclerosis Society (USA) Advisory Committee on Clinical Trials of New Agents in Multiple Sclerosis. Neurology 1996;46:907–911.

Madigand M, Oger JJ, Fauchet R, et al. HLA profiles in multiple sclerosis suggest two forms of disease and the existence of protective haplotypes. J Neurol Sci 1982;53:519–529.

Marrie RA, Wolfson C, Sturkenboom MC, et al. Multiple sclerosis and antecedent infections: a case–control study. Neurology 2000;54:2307–2310.

Marshall E. A shadow falls on hepatitis B vaccination effort. Science 1998;281:630–631.

Martin C, Enbom M, Soderstrom M, et al. Absence of seven human herpesviruses, including HHV-6, by polymerase chain reaction in CSF and blood from patients with multiple sclerosis and optic neuritis. Acta Neurol Scand 1997;95:280–283.

Martinez-Naves E, Victoria-Gutierrez M, Uria DF, Lopez-Larrea C. The germline repertoire of T cell receptor β-chain genes in multiple sclerosis patients from Spain. J Neuroimmunol 1993;47:9–13.

Martyn CN. The epidemiology of multiple sclerosis. In: Matthews WB, Compston A, Allen IV, Martyn CN, eds. McAlpine’s multiple Sclerosis. Edinburgh: Churchill-Livingstone, 1991, pp 2–40.

Martyn CN, Cruddas M, Compston DAS. Symptomatic Epstein-Barr virus infection and multiple sclerosis. J Neurol Neurosurg Psychiatry 1993;56:167–168.

Matthews B. Differential diagnosis of multiple sclerosis and related disorders. In: Compston A, ed. McAlpine’s Multiple Sclerosis, 3rd edition. New York: Churchill Livingstone, 1998a, p. 223–250.

(p.218) Matthews B. Symptoms and signs of multiple sclerosis. In: Compston A, ed. McAlpine’s Multiple Sclerosis, 3rd edition. New York: Churchill Livingstone, 1998b, pp 145–190.

Mayne M, Krishnan J, Metz L, et al. Infrequent detection of human herpesvirus 6 DNA in peripheral blood mononuclear cells from multiple sclerosis patients. Ann Neurol 1998;44:391–394.

McAlpine D. The benign form of MS. A study based on 241 cases seen within three years of onset and followed up until the tenth year or more of the disease. Brain 1961;84:186–203.

McDonald I. Diagnostic methods and investigation in multiple sclerosis. In: Compston A, ed. McAlpine’s Multiple Sclerosis, 3rd edition. New York: Churchill Livingstone, 1998, pp 251–279.

McDonald WI, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the Diagnosis of Multiple Sclerosis. Ann Neurol 2001;50:121–127.

Melbye M, Ebbesen P, Bennike T. Infectious mononucleosis in Greenland: a disease of the non-indigenous population. Scand J Infect Dis 1984;16:9–15.

Mickel HS. Multiple sclerosis: a new hypothesis. Perspect Biol Med 1975;18:363–374.

Midgard R, Albrektsen G, Riise T, et al. Prognostic factors for survival in multiple sclerosis: a longitudinal, population based study in More and Romsdal, Norway. J Neurol Neurosurg Psychiatry 1995;58:417–421.

Midgard R, Riise T, Nyland H. Epidemiologic trends in multiple sclerosis in More and Romsdal, Norway: a prevalence/incidence study in a stable population. Neurology 1991;41:887–892.

Midgard R, Riise T, Svanes C, et al. Incidence of multiple sclerosis in More and Romsdal, Norway from 1950 to 1991. An age-period-cohort analysis. Brain 1996;119:203–211.

Miller DH, Hammond SR, McLeod JG, et al. Multiple sclerosis in Australia and New Zealand: are the determinants genetic or environmental? J Neurol Neurosurg Psychiatry 1990;53:903–905.

Morré SA, De Groot CJ, Killestein J, et al. Is Chlamydia pneumoniae present in the central nervous system of multiple sclerosis patients? Ann Neurol 2000;48:399.

Morré SA, van Beek J, De Groot CJ, et al. Is Epstein-Barr virus present in the CNS of patients with MS? Neurology 2001;56:692.

Mumford CJ, Wood NW, Kellar-Wood H, et al. The British Isles survey of multiple sclerosis in twins. Neurology 1994;44:11–15.

Munch M, Møoller-Larsen A, Christensen T, et al. Production of retrovirus and Epstein-Barr virus in cell lines from multiple sclerosis patients. Acta Neurol Scand 1997;169(Suppl):65–69.

Munch M, Riisom K, Christensen T, et al. The significance of Epstein-Barr virus seropositivity in multiple sclerosis patients? Acta Neurol Scand 1998;97:171–174.

Munger KL, Peeling RW, Hernán M, et al. Infection with Chlamydia pneumoniae and risk of multiple sclerosis. Epidemiology 2003;14:141–147.

Myhr K-M, Riise T, Barrett-Connor E, et al. Altered antibody pattern to Epstein-Barr virus but not to other herpesviruses in multiple sclerosis: a population-based case–control study from western Norway. J Neurol Neurosurg Psychiatry 1998;64:539–542.

Myhr KM, Riise T, Vedeler C, et al. Disability and prognosis in multiple sclerosis: demographic and clinical variables important for the ability to walk and awarding of disability pension. Mult Scler 2001;7:59–65.

Nathanson N, Miller A. Epidemiology of multiple sclerosis: critique of the evidence for a viral etiology. Am J Epidemiol 1978;107:451–461.

Nelson LM, Hamman RF, Thompson DS, et al. Higher than expected prevalence of multiple sclerosis in northern Colorado: dependence on methodologic issues. Neuroepidemiology 1986;5:17–28.

Neutel CI, Walter SD, Mousseau G. Clustering during childhood of multiple sclerosis patients. J Chron Dis 1977;30:217–224.

Niederman JC, Evans AS. Epstein-Barr virus. In: Evans AS, Kaslow RA, eds. Viral Infections of Humans: Epidemiology and Control. New York: Plenum Medical Book Company, 1997, pp 253–283.

Nye FJ. Social class and infectious mononucleosis. J Hyg 1973;71:145–149.

Oksenberg JR, Gaiser CN, Cavalli-Sforza LL, Steinman L. Polymorphic markers of human T-cell receptor alpha and beta genes. Family studies and comparison of frequencies in healthy individuals and patients with multiple sclerosis and myasthenia gravis. Hum Immunol 1988;22:111–121.

Operskalski EA, Visscher BR, Malmgren RM, Detels R. A case–control study of multiple sclerosis. Neurology 1989;39:825–829.

Optic Neuritis Study Group. The 5-year risk of MS after optic neuritis: experience of the optic neuritis treatment trial. Neurology 1997;49:1404–1413.

O’Riordan JI, Losseff NA, Phatouros C, et al. Asymptomatic spinal cord lesions in clinically isolated optic nerve, brain stem, and spinal cord syndromes suggestive of demyelination. J Neurol Neurosurg Psychiatry 1998;64:353–357.

(p.219) Ormerod IE, Miller DH, McDonald WI, et al. The role of NMR imaging in the assessment of multiple sclerosis and isolated neurological lesions. A quantitative study. Brain 1987;110:1579–616.

Page WF, Kurtzke JF, Murphy FM, Norman JE Jr. Epidemiology of multiple sclerosis in U.S. veterans: V. Ancestry and the risk of multiple sclerosis. Ann Neurol 1993;33:632–639.

Peeling RW, Wang SP, Grayston JT, et al. Chlamydia pneumoniae serology: interlaboratory variation in microimmunofluorescence assay results. J Infect Dis 2000;181(Suppl 3):S426–S429.

Perkin GD, Bowden P, Rose FC. Smoking and optic neuritis. Postgrad Med J 1975;51:382–385.

Perkin GD, Rose FC. Uhthoff’s syndrome. Br J Ophthalmol 1976;60:60–63.

Phadke JG. Survival pattern and cause of death in patients with multiple sclerosis: results from an epidemiological survey in north east Scotland. J Neurol Neurosurg Psychiatry 1987;50:523–531.

Phadke JG. Clinical aspects of multiple sclerosis in north-east Scotland with particular reference to its course and prognosis. Brain 1990;113:1597–1628.

Poland SD, Rice GPA. Chlamydiae pneumonia and Multiple sclerosis. Neurology 2000:56:A165.

Poser CM. Trauma and multiple sclerosis. An hypothesis. J Neurol 1987;234:155–159.

Poser CM, Hibberd PL. Analysis of the ‘epidemic’ of multiple sclerosis in the Faroe Islands. II. Biostatistical aspects. Neuroepidemiology 1988;7:181–189.

Poser CM, Paty DW, Scheinberg L, et al. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol 1983;13:227–231.

Poser S, Kurtzke JF, Poser W, Schlaf G. Survival in multiple sclerosis. J Clin Epidemiol 1989;42:159–168.

Poser S, Poser W, Schlaf G, et al. Prognostic indicators in multiple sclerosis. Acta Neurol Scand 1986;74:387–392.

Poser S, Raun NE, Poser W. Age at onset, initial symptomatology and the course of multiple sclerosis. Acta Neurol Scand 1982;66:355–362.

Poser S, Ritter G, Bauer HJ, et al. HLA-antigens and the prognosis of multiple sclerosis. J Neurol 1981;225:219–221.

Poskanzer DC, Schapira K, Miller H. Multiple sclerosis and poliomyelitis. Lancet 1963;2:917–921.

Poskanzer DC, Sever JL, Sheridan JL, Prenney LB. Multiple sclerosis in the Orkney and Shetland Islands: IV. Viral antibody titres and viral infections. J Epidemiol Community Health 1980;34:258–264.

Poskanzer DC, Walker AM, Yonkondy J, Sheridan JL. Studies in the epidemiology of multiple sclerosis in the Orkney and Shetland Islands. Neurology 1976;Part 2:14–17.

Pucci E, Taus C, Cartechini E, et al. Lack of Chlamydia infection of the central nervous system in multiple sclerosis. Ann Neurol 2000;48:399–400.

Pugliatti M, Sotgiu S, Rosati G. The worldwide prevalence of multiple sclerosis. Clin Nerol Neurosurg 2002;104:182–191.

Rickinson AB, Kieff E. Epstein-Barr virus. In: Fields BN, Knipe DM, Howley PM, eds. Fields Virology. Philadelphia: Lippincott-Raven, 1996, pp 2397–2446.

Riise T. Cluster studies in multiple sclerosis. Neurology 1997;49(Suppl 2):S27–S32.

Riise T, Gronning M, Aarli JA, et al. Prognostic factors for life expectancy in multiple sclerosis analysed by Cox-models. J Clin Epidemiol 1988;41:1031–1036.

Riise T, Gronning M, Fernandez O, et al. Early prognostic factors for disability in multiple sclerosis, a European multicenter study. Acta Neurol Scand 1992;85:212–218.

Riise T, Grønning M, Klauber MR, et al. Clustering of residence of multiple sclerosis patients at age 13 to 20 years in Hordaland, Norway. Am J Epidemiol 1991;133:932–939.

Risch N. Linkage strategies for genetically complex traits. I. Multilocus models. Am J Hum Genet 1990;46:222–228.

Robertson N, Compston A. Surveying multiple sclerosis in the United Kingdom. J Neurol Neurosurg Psychiatry 1995;58:2–6.

Robertson NP, Fraser M, Deans J, et al. Age-adjusted recurrence risks for relatives of patients with multiple sclerosis. Brain 1996;119:449–455.

Robertson NP, O’Riordan JI, Chataway J, et al. Offspring recurrence rates and clinical characteristics of conjugal multiple sclerosis. Lancet 1997;349:1587–1590.

Rodriguez D, Delia Gaspera B, Zalc B, et al. Identification of a Val 145 Ile substitution in the human myelin oligodendrocyte glycoprotein: lack of association with multiple sclerosis. The Reseau de Recherche Clinique INSERM sur la Susceptibilite Genetique a la Sclerose en Plaques. Mult Scler 1997;3:377–381.

Rosati G. Descriptive epidemiology of multiple sclerosis in Europe in the 1980s: a critical overview. Ann Neurol 1994;36(S2):S164–S174.

(p.220) Rosati G. The prevalence of multiple sclerosis in the world: an update. Neurol Sci 2001;22:117–139.

Roth MP, Dolbois L, Borot N, et al. Myelin oligodendrocyte glycoprotein (MOG) gene polymorphisms and multiple sclerosis: no evidence of disease association with MOG. J Neuroimmunol 1995;61:117–122.

Roth MP, Nogueira L, Coppin H, et al. Tumor necrosis factor polymorphism in multiple sclerosis: no additional association independent of HLA. J Neuroimmunol 1994;51:93–99.

Runmarker B, Andersen O. Prognostic factors in a multiple sclerosis incidence cohort with twenty-five years of follow-up. Brain 1993;116:117–134.

Russell WR. Multiple sclerosis: occupation and social group at onset. Lancet 1971;2:832–834.

Saarela J, Schoenberg Fejzo MS, Chen D, et al. Fine mapping of a multiple sclerosis locus to 2.5 Mb on chromosome 17q22–q24. Hum Mol Genet 2002;11:2257–2267.

Sadovnick AD, Armstrong H, Rice GP, et al. A population-based study of multiple sclerosis in twins: update. Ann Neurol 1993;33:281–285.

Sadovnick AD, Baird PA, Ward RH. Multiple sclerosis: updated risks for relatives. Am J Med Genet 1988;29:533–541.

Sadovnick AD, Dyment D, Ebers GC. Genetic epidemiology of multiple sclerosis. Epidemiol Rev 1997;19:99–106.

Sadovnick AD, Ebers GC, Dyment DA, Risch NJ. Evidence for genetic basis of multiple sclerosis. The Canadian Collaborative Study Group. Lancet 1996;347:1728–1730.

Sandberg-Wollheim M, Ciusani E, Salmaggi A, Pociot F. An evaluation of tumor necrosis factor microsatellite alleles in genetic susceptibility to multiple sclerosis. Mult Scler 1995;1:181–185.

Sanders EA, Bollen EL, van der Velde EA. Presenting signs and symptoms in multiple sclerosis. Acta Neurol Scand 1986;73:269–272.

Sawcer S, Jones HB, Feakes R, et al. A genome screen in multiple sclerosis reveals susceptibility loci on chromosome 6p21 and 17q22. Nat Genet 1996;13:464–468.

Sherritt MA, Oksenberg J, de Rosbo NK, Bernard CC. Influence of HLA-DR2, HLA-DPw4, and T cell receptor alpha chain genes on the susceptibility to multiple sclerosis. Int Immunol 1992;4:177–181.

Shirodaria PV, Haire M, Fleming E, et al. Viral antibody titers. Comparison in patients with multiple sclerosis and rheumatoid arthritis. Arch Neurol 1987;44:1237–1241.

Silman AJ, Newman J, MacGregor AJ. Cigarette smoking increases the risk of rheumatoid arthritis. Results from a nationwide study of disease-discordant twins. Arthritis Rheum 1996;39:732–735.

Skegg DC, Corwin PA, Crave RS, et al. Occurrence of multiple sclerosis in the north and south of New Zealand. J Neurol Neurosurg Psychiatry 1987;50:134–139.

Smith ADM, Duckett S, Waters AH. Neuropathological changes in chronic cyanide intoxication. Nature 1963;200:179–181.

Soldan SS, Berti R, Salem N, et al. Association of human herpes virus 6 (HHV-6) with multiple sclerosis: increased IgM response to HHV-6 early antigen and detection of serum HHV-6 DNA. Nat Med 1997;3:1394–1397.

Soldan SS, Leist TP, Juhng KN, et al. Increased lymphoproliferative response to human herpesvirus type 6A variant in multiple sclerosis patients. Ann Neurol 2000;47:306–313.

Sopori ML, Kozak W. Immunomodulatory effects of cigarette smoke. J Neuroimmunol 1998;83:148–156.

Spillane J. The effect of nicotine on spinocerebellar ataxia. BMJ 1955;2:1345–1351.

Sriram S, Mitchell W, Stratton C. Multiple sclerosis associated with Chlamydia pneumoniae infection of the CNS. Neurology 1998;50:571–572.

Sriram S, Stratton CW, Yao S, et al. Chlamydia pneumonaie infection of the central nervous system in multiple sclerosis. Ann Neurol 1999;46:6–14.

Stazio A, Paddison RM, Kurland LT. Multiple sclerosis in New Orleans, Louisiana, and Winnipeg, Manitoba, Canada: follow-up of a previous survey in New Orleans, and comparison between the patient populations in the two communities. J Chron Dis 1967;20:311–332.

Stefferl A, Schubart A, Storch M, et al. Butyrophilin, a milk protein, modulates the encephalitogenic T cell response to myelin oligodendrocyte glycoprotein in experimental autoimmune encephalomyelitis. J Immunol 2000;165:2859–2865.

Stein EC, Schiffer RB, Hall WJ, Young N. Multiple sclerosis and the workplace: report of an industry-based cluster. Neurology 1987;37:1672–1677.

Sumaya CV, Myers LW, Ellison GW. Epstein-Barr virus antibodies in multiple sclerosis. Arch Neurol 1980;37:94–96.

Sumaya CV, Myers LW, Ellison GW, Ench Y. Increased prevalence and titer of Epstein-Barr virus antibodies in patients with multiple sclerosis. Ann Neurol 1985;17:371–377.

(p.221) Svenningsson A, Runmarker B, Lycke J, Andersen O. Incidence of MS during two fifteen-year periods in the Gothenburg region of Sweden. Acta Neurol Scand 1990;82:161–168.

Svenson LW, Woodhead SE, Platt GH. Regional variations in the prevalence rates of multiple sclerosis in the province of Alberta, Canada. Neuroepidemiology 1994;13:8–13.

Swank RL, Lerstad O, Strom A, Backer J. Multiple sclerosis in rural Norway. N Engl J Med 1952;246:721–728.

Swingler RJ, Compston D. The distribution of multiple sclerosis in the United Kingdom. J Neurol Neurosurg Psychiatry 1986;49:1115–1124.

Taus C, Pucci E, Cartechini E, et al. Absence of HHV-6 and HHV-7 in cerebrospinal fluid in relapsing-remitting multiple sclerosis. Acta Neurol Scand 2000;101:224–228.

Thompson AJ, Hutchinson M, Brazil J, et al. A clinical and laboratory study of benign multiple sclerosis. Q J Med 1986;58:69–80.

Thorogood M, Hannaford PC. The influence of oral contraceptives on the risk of mulitple sclerosis. Br J Obstet Gynaecol 1998;105:1296–1299.

Tompkins LS, Schachter J, Boman J, et al. Collaborative multidisciplinary workshop report: detection, culture, serology, and antimicrobial susceptibility testing of Chlamydia pneumoniae. J Infect Dis 2000;181(Suppl 3):S460–S461.

Touze E, Gout O, Verdier-Taillefer MH, et al. The first episode of central nervous system demyelinization and hepatitis B virus vaccination. Rev Neurol (Paris) 2000;156:242–246.

Vassallo L, Elian M, Dean G. Multiple sclerosis in southern Europe. II: Prevalence in Malta in 1978. J Epidemiol Community Health 1979;33:111–113.

Vaughan JH, Riise T, Rhodes GH, et al. An Epstein-Barr virus–related cross reactive autoimmune response in multiple sclerosis in Norway. J Neuroimmunol 1996;69:95–102.

Verjans E, Theys P, Delmotte P, Carton H. Clinical parameters and intrathecal IgG synthesis as prognostic features in multiple sclerosis. Part I. J Neurol 1983;229:155–165.

Vessey M, Doll R, Peto R, et al. A long-term follow-up study of women using different methods of contraception—an interim report. J Biosoc Sci 1976;8:373–427.

Vessey MP, Villard-Mackintosh L, Yeates D. Oral contraceptives, cigarette smoking and other factors in relation to arthritis. Contraception 1987;35:457–464.

Villard-Mackintosh L, Vessey MP. Oral contraceptives and reproductive factors in multiple sclerosis incidence. Contraception 1993;47:161–168.

Visscher BR, Detels R, Coulson AH, et al. Latitude, migration, and the prevalence of multiple sclerosis. Am J Epidemiol 1977;106:470–475.

Visscher BR, Liu KS, Clark VA, Detels R, et al. Onset symptoms as predictors of mortality and disability in multiple sclerosis. Acta Neurol Scand 1984;70:321–328.

Voigt LF, Koepsell TD, Nelson JL, et al. Smoking, obesity, alcohol consumption, and the risk of rheumatoid arthritis. Epidemiology 1994;5:525–532.

Wallin MT, Page WF, Kurtzke JF. Epidemiology of multiple sclerosis in US veterans. VIII. Long-term survival after onset of multiple sclerosis. Brain 2000;123:1677–1687.

Wandinger K, Jabs W, Siekhaus A, et al. Association between clinical disease activity and Epstein-Barr virus reactivation in MS. Neurology 2000;55:178–184.

Warner HB, Carp RI. Multiple sclerosis and Epstein-Barr virus. Lancet 1981;2:1290.

Warren SA, Warren KG, Greenhill S, Paterson M. How multiple sclerosis is related to animal illness, stress and diabetes. Can Med Assoc J 1982;126:377–385.

Weinshenker BG, Rice GP, Noseworthy JH, et al. The natural history of multiple sclerosis: a geographically based study. 3. Multivariate analysis of predictive factors and models of outcome. Brain 1991;114:1045–1056.

Whitacre CC. Sex differences in autoimmune disease: focus on multiple sclerosis. Nat Immunol 2001;2:777–780.

Whitacre CC, Reingold SC, O’Looney PA. A gender gap in autoimmunity. Science 1999;283:1277–1278.

Wilborn F, Schmidt CA, Brinkmann V, et al. A potential role for human herpesvirus type 6 in nervous sytem disease. J Neuroimmunol 1994;49:213–214.

Willett WC. Nutritional Epidemiology, 2nd edition. New York: Oxford University Press, 1998.

Winer S, Astsaturov I, Cheung RK, et al. T cells of multiple sclerosis patients target a common environmental peptide that causes encephalitis in mice. J Immunol 2001;166:4751–4756.

Wingerchuk D, Liu Q, Sobell J, et al. A population-based case–control study of the tumor necrosis factor alpha-308 polymorphism in multiple sclerosis. Neurology 1997;49:626–628.

Wucherpfennig KW, Strominger JL. Molecular mimicry in T cell-mediated autoimmunity: viral peptides activate human T cell clones specific for myelin basic protein. Cell 1995;80:695–705.

(p.222) Wynn DR, Rodriguez M, O’Fallon WM, Kurland LT. A reappraisal of the epidemiology of multiple sclerosis in Olmsted County, Minnesota. Neurology 1990;40:780–786.

Zhang S, Hernán M, Olek M, et al. Intakes of carotenoids, vitamin C, and vitamin E and MS risk among two large cohorts of women. Neurology 2001;57:75–80.

Zhang SM, Willett WC, Hernan MA, et al. Dietary fat in relation to risk of multiple sclerosis among two large cohorts of women. Am J Epidemiol 2000;152:1056–1064.