key: cord-023747-mvq6353a
authors: Ascherio, Alberto; Munger, Kassandra L.
title: Epidemiology of Multiple Sclerosis: Environmental Factors
date: 2009-12-25
journal: nan
DOI: 10.1016/b978-1-4160-6068-0.00004-8
sha: 
doc_id: 23747
cord_uid: mvq6353a

This chapter discusses the environmental factors associated to epidemiology of multiple sclerosis. The epidemiologic evidence points to three environ­mental risk factors—infection with the Epstein-Barr virus (EBV), low levels of vitamin D, and cigarette smoking—whose association with multiple sclerosis (MS) seems to satisfy in varying degrees most of the criteria that support causality, including temporality, strength, consis­tency, biologic gradient, and plausibility. None of these associations, however, has been tested experimentally in humans and only one––vitamin D deficiency is presently amenable to experimental interventions. The evidence, albeit more sparse and inconsistent, linking other environmental factors to MS risk are summarized. Epidemiologic clues to the hypothetical role of infection in MS are com­plex and often seem to point in opposite directions. The ecological studies, database/linkage analyses, and longitudinal studies of sunlight exposure and vitamin D are reviewed. Biologic mechanisms for smoking and increased risk of MS could be neuro­toxic, immunomodulatory, vascular, or they could involve increased frequency and duration of respiratory infections. Some other possible risk factors include––diet and hepatitis B vaccine.

data. Such has been the case, for example, with interventions to reduce lung cancer incidence by reducing exposure to tobacco smoke.

As discussed in this chapter, epidemiologic evidence points to three environmental risk factors-infection with the Epstein-Barr virus (EBV), low levels of vitamin D, and cigarette smoking-whose association with multiple sclerosis (MS) seems to satisfy in varying degrees most of the criteria that support causality, including temporality (i.e., the cause must precede the effect), strength, consistency, biologic gradient, and plausibility. None of these associations, however, has been tested experimentally in humans, and only one (vitamin D deficiency) is presently amenable to experimental interventions. This chapter also summarizes the evidence, albeit more sparse and inconsistent, linking other environmental factors to MS risk.

For many years, it appeared that the "who, where, and when" of MS epidemiology was well understood. However, some aspects of MS epidemiology may be changing, notably the observations of an attenuation of the latitude gradient 3, 4 and the increasing female-to-male ratio. 5 In this section, we discuss the "classic" view of MS epidemiology, some of which has been known for more than 50 years, and then some recent developments that may provide new clues to the etiology of MS.

MS is the most common neurologic disease in young adults. Incidence rates are low in childhood and adolescence (<6/100,000/year) high in the middle to late twenties and early thirties (11 to 18/100,000/year in high-risk populations), and gradually decline thereafter, with rates less than 9/100,000/year among those older than 45 years of age. 3, 6 Women are approximately twice as likely as men to develop MS, 7, 8 and the lifetime risk among white women is about 1 in 200. 3, 9 MS exhibits a worldwide latitude gradient, with high prevalence and incidence in Northern Europe, 7 Canada, 10,11 the northern United States, 3, 12, 13 and southern Australia 14 and decreasing prevalence and incidence in regions closer to the equator. 15 Exceptions to the latitude gradient exist and include a lower than expected prevalence in Japan 16 and higher than expected prevalence and incidence in the Mediterranean islands of Sardinia and Sicily. 7 Kurtzke 17 summarized the early descriptive studies by depicting areas of high (≥30/100,000), medium (5 to 29/100,000), and low (<5/100,000) prevalence of MS; we have updated his figures with more recent prevalence estimates 7, 16, [18] [19] [20] [21] [22] [23] [24] (Fig. 4-1) . A more comprehensive review of MS incidence and prevalence worldwide was published in 2005. 18 It is important to note that differences in estimated incidence across countries or time periods can result from differences in study design, case ascertainment, or diagnostic criteria, rather than from real changes in disease occurrence. Differences in prevalence are even more difficult to interpret, because they may reflect increased survival or earlier diagnosis, both of which can occur even if the incidence is the same. 25 In spite of these limitations, the collective data do support a higher risk of MS at higher latitudes, both north and south of the equator.

The existence of the latitude gradient alone is not enough to support an environmental component, because it could be explained by genetic differences. 18, 26 However, studies of MS incidence and prevalence among migrant populations also support a role for environmental factors. These studies have limitations, in that migrants may be different from nonmigrants in socioeconomic and health status, may not utilize local health care resources, and therefore they may be less likely to be diagnosed; in addition, enumeration of the immigrant population for disease statistics may be difficult or impossible. 25, 27 Nevertheless, migrant studies on MS collectively support a decreased prevalence of MS among those who migrate from high-to low-risk areas, particularly if the migration occurs before 15 years of age. 27 Moreover, one study found a decreased prevalence of MS in all age groups among immigrants from Europe to Australia, suggesting that the protective effect may extend into adulthood as well. 28 Studies within the United States have also supported a decreased risk of MS among migrants from northern (>41° to 42° N), Australia and New Zealand Europe Figure 4-1 Worldwide prevalence estimates of multiple sclerosis. Blue, more than 90 cases per 100,000 population; purple, 60 to 89/100,000; green, 30 to 59/100,000; orange, 5 to 29/100,000; yellow, fewer than 5/100,000; white, insufficient data. An asterisk indicates that data for that region or country are older and should be interpreted cautiously.

high-risk parts of the country to southern (<37° N), low-risk regions. 29, 30 The study of U.S. veterans 30 is particularly compelling because of its large sample size and rigorous design. In this study, Kurtzke observed that individuals who were born in the northern United States but migrated to the southern part of the country before joining the military had a 50% reduced risk of MS compared with those who did not migrate ( Fig. 4-2) .

Fewer studies have been conducted among migrants from low-to high-risk areas. In general, these studies have found that a low risk of MS is retained after migration, but that the offspring of migrants have a higher risk of MS, similar to that in the host country. 27, [31] [32] [33] [34] [35] In the U.S. veterans study, 30 individuals who were born in the southern part of the country and migrated to northern states before entering the military had a 20% increased risk of MS, and those migrating from the middle tier of states to northern regions had a 31% increased risk (see Fig. 4 -2). More recently, in a study conducted in the French West Indies (a low-risk area), an increased risk of MS was found among individuals who had moved to France (a high-risk area) and then returned to the West Indies. The increase in risk was greatest among those who migrated to France before the age of 15 years. 36

The incidence of MS appears to have been relatively stable over the past 50 years in several high-risk areas, including Denmark 6 and the northern United States, 37 but there is some evidence that MS may be increasing in Japan 16 and in parts of Southern Europe, most notably in Sardinia. 38 Interestingly, the island of Malta has continued to experience low, stable rates of MS 39 despite its proximity to Sardinia and Sicily and a high frequency of the MS-associated HLA-DRB1*1501 allele. 40 There is also evidence of an increased female-to-male ratio in MS incidence. In Canada, the female-to-male ratio apparently has increased from approximately 2:1 among individuals born in the 1930s and 1940s to approximately 3:1 among those born in the 1970s. 5 This change is strongly correlated with, and could be at least in part explained by, a sharp increase in the female-to-male ratio in smoking behavior (unpublished data), because smoking is a strong risk factor for MS (see later discussion).

An attenuation of the latitude gradient was observed independently in a population of U.S. nurses 3 and in U.S. military veterans. 4 Among nurses born between 1920 and 1946 and among veterans of World War II or the Korean conflict, those living in the northern tier of states (>41° to 42° N) had a greater than threefold increased risk of MS compared to those in the southern tier (<37° N). Among Vietnam and Gulf War veterans, however, this gradient was attenuated to less than twofold, and among nurses born between 1947 and 1964 it completely disappeared ( Fig. 4-3) . Because the methods used to determine rates of MS in the early and later cohorts were the same, and because the individuals in the cohorts had similar socioeconomic status 3 or access to health care, 4 this attenuation was unlikely to be due to artifact. A change of this magnitude over such a short period of time argues for an environmental, rather than a genetic, explanation of the latitude gradient; as discussed later, this environmental factor may involve changes in patterns of infection or sun exposure, or both. Further, the attenuation was probably caused by an increase in MS incidence in the southern United States, because incidence rates in the northern states, based at least on data from the longitudinal study in Olmsted County, Minnesota, seem to have remained relatively stable. 37 An attenuation of the latitude gradient in Europe has also been observed; however, no systematic studies have assessed this gradient within the same population over time, and the attenuation therefore may be due to improved study methodology and case ascertainment, particularly within the United Kingdom. 18

The possibility of an infectious cause was considered early in MS history, and numerous viruses and bacteria were, at different times, implicated as likely etiologic agents. The results of early studies, based on microscopic examination of pathologic material and attempts to transmit the disease to animals, often were null or spuriously positive because of contamination and could not be replicated.

Later, numerous serologic studies were conducted, often demonstrating significantly elevated antibody titers against several viruses in MS patients compared with healthy controls, but these differences were probably an epiphenomenon of the immune activation rather than being of etiologic significance. 41 In part as a consequence of these investigations, many researchers became skeptical about the existence of an infectious agent causing MS, and this skepticism persists today. Epidemiologic clues to the hypothetical role of infection in MS are complex and often seem to point in opposite directions. On the one hand, results of family studies, including investigations of half-siblings, adopted children, and spouses of individuals with MS, support a strong genetic component as the leading explanation of MS clustering within families and provide little evidence of person-to-person transmission. 42 On the other, there are well-documented, albeit controversial, 43 reports of epidemics of MS, most notably in the Faroe Islands, 44 that are most easily explained by the introduction and transmission of an infectious agent. To reconcile these findings, it has been postulated that MS is a rare complication of a common infection, with the disease occurring in genetically or otherwise predisposed individuals. In this scenario, the epidemics would be a consequence of the introduction of the MS-causing agent for the first time in remote, previously naïve populations. 45 Two hypotheses as to the nature of this infection have been proposed: (1) the responsible microorganism is more common in areas of high MS prevalence (the "prevalence" hypothesis), and (2) the MS-causing agent is ubiquitous and more easily transmitted in areas of low MS prevalence, where infection occurs predominantly in infancy, when it would be less harmful and more likely to confer protective immunity. The latter proposal is called the "poliomyelitis" hypothesis, by analogy with the epidemiology of poliomyelitis before vaccination. 46 The poliomyelitis hypothesis is also consistent with the higher prevalence of MS in communities with better hygiene, 47 in individuals with higher education, 48, 49 and in those with late age at infection with common viruses, 50 as well as the general lack of increase in MS incidence among individuals migrating from low-to high-prevalence areas. 27 However, the poliomyelitis hypothesis cannot explain the reduced risk of MS among migrants from high-to low-risk areas and, in fact, would predict an increase in MS risk in this circumstance, whereas the prevalence hypothesis is consistent with the observations. Failure to identify a specific microbe as the cause of MS, despite evidence that is consistent with some role for infection in at least modulating MS risk, has strengthened support for a third, more general, "hygiene" hypothesis, according to which exposure to multiple infections in childhood primes the immune responses later in life toward a less inflammatory and a less autoimmunogenic profile. 51 The hygiene hypothesis can explain all the features of MS epidemiology that are explained by the original formulation of the poliomyelitis hypothesis. In addition, the protective effect of migration from high-to low-MS areas, which is paradoxical under the poliomyelitis hypothesis, could be beneficial because of increased exposure of migrants to parasitic and other infections in the low-risk area. At the population level, prevalence of MS is positively correlated with high levels of hygiene, as measured, for example, by prevalence of intestinal parasites. 52 The improving hygienic conditions in southern Europe in the last few decades could explain the increased prevalence of MS reported in multiple surveys (although whether there was a true increase in MS incidence remains unsettled). 7 It is also interesting that infection with intestinal helminths, which is highly prevalent in developing countries, had been reported to cause an immune deviation with attenuation of helper T-cell 1 cellular immune responses and remission of MS. 53 Finally, the hygiene hypothesis provides a convincing explanation for the observations that infectious mononucleosis (IM) is associated with an increased risk of MS (relative risk [RR] = 2.3; P < .00000001) 54 and that the epidemiology of IM is strikingly similar to that of MS (Table 4 -1). 55 Because IM is common in individuals who are first infected with EBV in adolescence or adulthood 56 but rare when EBV infection occurs in childhood, it is a strong marker of age at EBV infection, which is itself strongly correlated with socioeconomic development across populations and with socioeconomic status within populations. 57 An exception to this pattern is seen in Asia, where EBV infection occurs uniformly early in life and IM is thus rare. It is noteworthy that the incidence of MS remains relatively low in Asian countries, including Japan, despite the fast industrialization and reduction of infectious diseases, 58 although there is evidence that the incidence may be increasing in Japan. 16 According to the hygiene hypothesis, the association between IM and MS risk does not reflect a causal effect of EBV but rather the indirect manifestation of a common cause; that is, both MS and IM are the result of high hygiene and a resulting low burden of infection during childhood. An important prediction of this hypothesis is that MS risk will be high among individuals reared in a highly hygienic environment, even if they do not happen to be infected with EBV later in life, whereas, if EBV has a causal role in MS, individuals who are not infected with EBV would have a low risk of MS. 59 The data on this point are unequivocal: individuals who are not infected with EBV, even though they have the same hygienic upbringing as those with IM, have an extremely low risk of MS (odds ratio [OR] from metaanalysis = 0.06; P < .00000001) ( Table 4 -2). The contrast could not be sharper or more consistent: MS risk among individuals who are not infected with EBV is at least 10-fold lower than that of individuals who are EBV-positive, and 20-fold lower than that of individuals with a history of IM. 59 Because studies in pediatric MS 60,61 rule out a common genetic resistance to MS and EBV infection, 59 we can conclude either that EBV itself or some other factor closely related to EBV is a strong causal risk factor for MS or that MS itself strongly predisposes to EBV infection.

Temporality is the only truly necessary criteria for causality. The association between EBV infection and MS is strong and consistent across multiple studies in different populations, and there is to some extent a biologic gradient (higher risk associated with severity of infection, as indicated by history of IM). Until recently, all studies on MS and infection used a cross-sectional design and could not completely rule out the possibility that EBV infection was a consequence rather than a cause of MS. However, the results of four longitudinal serologic studies have now been published (Table 4-3) . [62] [63] [64] [65] The most consistent finding across these studies is that, among individuals who will develop MS, there is an elevation of serum antibodies against the EBV nuclear antigen 1 (EBNA1) that precedes the onset of MS symptoms by many years. The presence of anti-EBNA1 antibodies is a marker of past infection with EBV, because titers typically rise only weeks after the acute infection. Further, there is no evidence in clinical studies of acute primary EBV infection in individuals with MS. 66 Taken together, these results indicate that MS is a consequence rather than a cause of EBV infection.

Until recently, EBV had not been found in MS lesions, 67, 68 and therefore the link between EBV and MS was postulated to be mediated by indirect mechanisms. The leading hypothesis was that the immune response to EBV infection in genetically susceptible individuals cross-reacts with myelin antigens (molecular mimicry). The discovery that MS patients have an increased frequency and broadened specificity of CD4-positive T cells recognizing EBNA1 69 and the identification of two EBV peptides (one of which is from EBNA1) as targets of the immune response in the cerebrospinal fluid of MS patients 70 provided support to the molecular mimicry theory. Other proposed hypotheses included the activation of superantigens, 71 an increased expression of alpha B-crystallin, 72 and infection of autoreactive B lymphocytes. 73 However, in a recent, rigorous pathologic study, 74 large numbers of EBVinfected B cells were found in the brain of most of MS patients. These cells were more numerous in areas with active inflammatory infiltrates, where cytotoxic CD8-positive T cells displaying an activated phenotype were seen contiguous to the EBV-infected cells. Alone, these pathologic findings provide only suggestive evidence for a causal role of EBV in MS, because the infiltration of EBV-infected B cells could be secondary to the inflammatory process that is the hallmark of MS, but their convergence with the epidemiologic evidence described earlier 59 is so striking that noncausal explanations become improbable. However, independent replication of these findings is needed before any conclusion can be drawn. The strong increase in MS risk after EBV infection and (if confirmed) the presence of EBV in MS lesions suggest that antiviral drugs or a vaccine against EBV could contribute to MS treatment and prevention. Although antiviral drugs have been tried in the past for MS treatment with borderline results, 75-77 none of the treatment regimens used was sufficiently effective against latent EBV infection.

Several aspects of MS epidemiology cannot be explained by EBV infection, indicating that other factors must contribute. 59 Genes are clearly important, and it is of interest that the association between anti-EBNA1 titers and MS risk has been found in both HLA-DRB1*1501-positive and HLA-DRB1*1501-negative individuals. 78 Variations in EBV strains could also play a role, although evidence in support of this hypothesis remains limited. 79, 80 Many other infectious agents have been hypothesized to be related to MS, mostly because of pathologic studies or their role in animal models. Recent candidates include Chlamydia pneumoniae, 81-84 human herpesvirus 6, 85-87 retroviruses, 88, 89 and coronaviruses, 90 but there are no convincing epidemiologic studies linking these infections to MS risk. Noninfectious factors may also be important, and prominent among them are vitamin D and cigarette smoking.

One of the strongest correlates of latitude is the duration and intensity of sunlight, which in ecologic studies is inversely correlated with MS prevalence. [92] [93] [94] Because exposure to sunlight is for most people the major source of vitamin D, 95 average levels of vitamin D also display a strong latitude gradient. Ultraviolet B (UV-B) radiation (290 to 320 nm) converts cutaneous 7-dehydrocholesterol to previtamin D 3 . Previtamin D 3 spontaneously isomerizes to vitamin D 3 , which is then hydroxylated to 25(OH)D 3 (25-hydroxyvitamin D 3 ), the main circulating form of the vitamin, and then to 1,25(OH) 2 D 3 (1,25-dihydroxyvitamin D 3 ), the biologically active hormone. 95 However, during the winter months at latitudes greater than 42° N (e.g., Boston, MA), even prolonged sun exposure is insufficient to generate vitamin D, 96 and levels decline. 97, 98 Use of supplements or high consumption of fatty fish (a good source of vitamin D) or vitamin D-fortified foods (mostly milk in the United States) may partially compensate for this decline, but few people consume large enough amounts of vitamin D, and seasonal deficiency is common. A link between vitamin D deficiency and MS was proposed more than 30 years ago as a possible explanation of the latitude gradient and of the lower prevalence of MS in fishing communities with high levels of fish intake 99 ; however, the immunomodulatory effects of vitamin D were not known, and the hypothesis did not generate much interest at the time. After the discovery that the vitamin D receptor is expressed in several cells in the immune system and is a potent immunomodulator, 100 a series of experiments revealed a protective role of 1,25(OH) 2 D 3 in several autoimmune conditions and in transplant rejection. 100 The effects in experimental autoimmune encephalomyelitis, an animal model of MS, were particularly striking: injection of 1,25(OH) 2 D 3 was found to completely prevent the clinical and pathologic signs of disease, 101,102 whereas vitamin D deficiency accelerated the disease onset. 102, 103 With vitamin D deficiency becoming a biologically plausible risk factor for MS, several epidemiologic studies were conducted to determine whether exposure to sunlight or vitamin D intake is associated with MS risk. The main results of these studies are shown in Table 4 -4, and their strengths and limitations are discussed in the following paragraphs.

As mentioned earlier, the results of ecologic studies support an inverse association between sunlight exposure and MS risk. However, because people living in the same area share many characteristics other than the level of sunlight, the consensus is that evidence from these studies is weak.

In an exploratory investigation based on death certificates, working outdoors was associated with a significantly lower MS mortality in areas of high, but not low, sunlight. 104 In a separate study in the United Kingdom, the skin cancer rate, 

a marker of sunlight exposure, was found to be about 50% lower than expected among individuals with MS (P = .03). 106 Although the results of these investigations are consistent with a protective effect of UV light exposure, they could also represent "reverse causation" (i.e., individuals with MS could reduce their exposure to sunlight after disease onset).

The results of case-control studies comparing history of sun exposure in childhood (presumed to be a critical period, mostly from the results of studies in migrants) between MS cases and controls have been conflicting. The results of one study were contrary to a protective effect of vitamin D, 106 and no association between sun exposure in childhood and MS risk was found in another. 107 In contrast, results consistent with a protective effect of sun exposure were reported in a study in Tasmania in which information on time spent in the sun was complemented by measurement of skin actinic damage, a biomarker of UV light exposure, 108 as well as an investigation in Norway 109 and a study of monozygotic twins in the United States. 110 In the Norway study, an inverse association was also found between consumption of fish and MS risk. Selection and recall biases are potential problems in case-control studies, but recall bias cannot explain the inverse association observed in Tasmania with actinic damage, 108 and selection bias is unlikely in the twin study.

The strongest evidence relating vitamin D levels to MS risk has been provided by two longitudinal studies, one based on assessment of dietary vitamin D intake, and one on serum levels of 25(OH)D. The relation between vitamin D intake and MS risk was studied in more than 200,000 women in the Nurses' Health Study and Nurses' Health Study II cohorts. 111 Dietary vitamin D intake was assessed from comprehensive and previously validated semiquantitative food frequency questionnaires administered every 4 years during the follow-up of the cohorts. 112, 113 Total vitamin D intake at baseline was inversely associated with risk of MS: the age-adjusted pooled relative risk (RR) comparing the highest with the lowest quintile of consumption was 0.67 (95% confidence interval [CI], 0.40 to 1.12; P for trend = .03). Intake of 400 IU/day of vitamin D from supplements only was associated with a 40% lower risk of MS. These RRs did not materially change after further adjustment for pack-years of smoking and latitude at birth. Confounding by other micronutrients cannot be excluded, but adjustments for them in the analyses did not change the results. Because dietary vitamin D is only one component contributing to total vitamin D status (the other being sun exposure), a determination of whether serum levels of vitamin D are associated with MS risk in healthy individuals would strengthen the evidence in favor of a causal role for vitamin D. The serum level of 25(OH)D is a marker of vitamin D status and bioavailability; therefore, if vitamin D is protective, high serum levels of 25(OH)D would be expected to predict a lower risk of MS in healthy individuals. This question was recently addressed in a collaborative, prospective case-control study using the Department of Defense Serum Repository (DoDSR). 114 The study included 257 military personnel with confirmed MS and at least two serum samples collected before the onset of MS symptoms. Risk of MS was 51% lower among white individuals with 25(OH)D levels of 100 nmol/L or higher, compared with those levels lower than 75 nmol/L, and the reduction in MS risk associated with 25(OH)D levels � 100 nmol/L compared with those levels < 100nmol/L was considerably stronger before the age of 20 years (16 to 19 years) than at ages 20 or older.

An important question concerning vitamin D and MS is the age intervals during which vitamin D may be important. The results of migration studies suggest that more pronounced changes in MS risk are likely to occur among individuals who migrate in childhood. The age of 15 years, chosen as an arbitrary cutoff point in early studies, is usually quoted in the literature, but the reality is that data are insufficient to identify a meaningful threshold above which migration would not alter MS risk, 27 and in at least one study a reduction in risk was also observed among individuals who migrated as adults. 115 The results of the casecontrol study in Tasmania suggest that exposure to sunlight is mostly protective in childhood. 108 Further, vitamin D exposure in utero has been proposed as a possible explanation for the peak in MS incidence among individuals born in May (whose mothers were not pregnant during the summer, when UV light levels are higher) and the dip among those born in November, according to recent data from Canada and Sweden. 116 On the other hand, the results of the longitudinal studies support a protective effect of vitamin D also later in life. Both the lower risk of MS among women taking vitamin D supplements 111 and the lower risk among men and women with higher levels of 25(OH)D 114 would be difficult to explain by a protective effect of vitamin D solely in utero or during childhood. Therefore, it seems likely that, if vitamin D effectively protects against MS, levels during early adult life are also important.

Overall, the epidemiologic evidence of a causal association between vitamin D and MS is strong but not compelling, mainly because there are few studies based on prospective measurement of levels of exposure to sunlight, vitamin D intake, or serum 25(OH)D concentration. However, the public health implications of a possible causal association are enormous. If vitamin D reduces the risk of MS, supplementation in adolescents and young adults could be used effectively for prevention. Based on studies among individuals with low sun exposure, supplements providing between 1000 and 4000 IU/day of vitamin D would increase serum 25(OH)D to the optimal levels. [117] [118] [119] [120] There is an urgent need to conduct further longitudinal studies, preferably in a large, randomized controlled clinical trial assessing whether vitamin D supplementation in the general population prevents MS. The trial would have to be very large, because MS is a rare disease, but the sample size could be reduced by oversampling individuals who are at high risk, such as those with first-degree relatives who have MS. Alternative study designs might include national or multinational studies based on randomization of school districts or other suitable units.

Cigarette smoking was found to increase the risk of MS in some 121, 122 but not all 123, 124 case-control studies. A cross-sectional survey of the general population in Hordaland County, Norway, found an increased risk of MS in ever-smokers compared with never-smokers (RR = 1.8; 95% CI, 1.1 to 2.9). 125 Four prospective studies on smoking and MS have been conducted. Among 17,000 British women in the Oxford Family Planning Association Study, those who smoked 15 or more cigarettes per day were compared with never-smokers and had an 80% increased risk of MS (RR = 1.8; 95% CI, 0.8 to 3.6). 126 A total of 46,000 women from across the United Kingdom were enrolled in the Royal College of General Practitioners' Oral Contraception Study, which found that women smoking 15 or more cigarettes per day had a 40% increased risk of MS (RR = 1.4; 95% CI, 0.9 to 2.2), compared with never-smokers. 127 The Nurses' Health Study and Nurses' Health Study II cohorts included more than 200,000 U.S. women; those who smoked 25 or more pack-years had a 70% increased risk (RR = 1.7; 95% CI, 1.2 to 2.4; P < .01) compared with never-smokers. 128 In a prospective case-control study in the General Practice Research Database, which included both men and women, ever-smokers had a 30% increased risk of MS, compared with never-smokers (RR = 1.3; 95% CI, 1.0 to 1.7). 129 The suggestion of an increased risk of MS among smokers was consistent across all four studies, and pooled estimates of the relative risk were highly statistically significant when never-smokers were compared with past and current smokers (Fig. 4 -4A) or with moderate and heavy smokers (see Fig. 4 -4B). Additional support for a role of smoking includes a twofold increase in risk of pediatric MS among children exposed to parental smoking 130 and an increased risk of transition to secondary progressive MS among individuals with relapsing-remitting MS 129 ; however, the latter finding was not confirmed in a recent investigation. 131 Biologic mechanisms for smoking and increased risk of MS could be neurotoxic, 132 immunomodulatory, 133, 134 or vascular (i.e., increased permeability of the blood-brain barrier), or they could involve increased frequency and duration of respiratory infections, 135 which may then contribute to increased MS risk. Smoking also appears to increase the risk of other autoimmune diseases, including rheumatoid arthritis [136] [137] [138] [139] [140] and systemic lupus erythematosus, 141 arguing for a more general effect of cigarette smoking on autoimmunity.

Although several foods or nutrients were found to be related to be MS risk in ecologic or case-control studies, the results overall were inconsistent and unconvincing. In ecologic studies, positive correlations were found between MS and intake of animal fat [142] [143] [144] and saturated fat, 144 as well as consumption of meat, 145 milk, and butter, 143, 146 and inverse correlations were found with intake of fat from fish 143, 145 and nuts 143 (sources of polyunsaturated fat). An increased risk of MS with increasing animal or saturated fat intake and a protective effect of increasing polyunsaturated fat intake were also reported in a case-control study, 147 but otherwise the results of case-control studies have largely not supported an association between increased MS risk and milk or meat consumption, 121, [147] [148] [149] [150] or decreased risk and consumption of sources of polyunsaturated fat such as fish or nuts. 121, 147 However, in a recent study in Norway, 109 fish consumption 3 or more times per week among individuals living at latitudes between 66° and 71° N was inversely related to MS risk. Other results have included an inverse association of risk with intake of vitamin C and juice, 147 but no association with other antioxidant vitamins 147 or with fruits and vegetables 121, 123, 147, 151 has been reported.

It is important to note that ecologic studies are prone to be confounding and in general provide only very weak evidence of the potential effects of diet on disease risk. Retrospective case-control studies are also prone to bias due to both control selection and differential recall. The latter effect is particularly problematic, because even a modest difference in diet recall between cases and controls can cause a large bias in relative risk estimates. 152 This problem is compounded in MS by changes in diet that may occur in the early clinical or preclinical phases of the disease. Therefore, although these studies have been important in drawing attention to several aspects of diet as potentially important risk factors for MS, their results, whether in favor or against an hypothetical association, should be interpreted extremely cautiously. Understanding of the relation between diet and MS will require the conduct of large longitudinal investigations, with repeated assessment of diet using rigorous and validated methods and possibly measurements of biomarkers of nutrient intakes. So far, the only prospective studies of diet and MS were those conducted among women in the two Nurses' Health Study cohorts. In this population, neither animal fat nor saturated fat was associated with MS risk, but there was a suggestion of an inverse association with intake of the n-3 polyunsaturated fat linolenic acid. 153 There were also no significant associations between intakes of dairy products, fish, meat, 153 vitamins C or E, carotenoids, or fruits and vegetables and MS risk. 154 However, participants in these studies were already 25 to 55 years of age at time of recruitment, and therefore they shed little light on the possible effect of diet earlier in life and MS risk.

Studies have also examined whether intake of polyunsaturated fats affects MS progression. N-3 polyunsaturated fat supplementation in doses ranging from 2.85 to 3.90 g/day administered for periods of 6 to 24 months did not have significant effects on disability levels in two randomized controlled trials that included a total of 339 patients with relapsing-remitting MS, 155, 156 although trends were in favor of the supplemented groups in both studies. Results of three randomized controlled trials examining the effects of n-6 polyunsaturated fat supplementation (17 to 20 g/day for 24 to 30 months) on MS progression, including a total of 279 patients with relapsing-remitting MS, [157] [158] [159] and a meta-analysis of these studies 160 suggested that supplementation may reduce the severity and duration of relapses.

In summary, there is no compelling evidence that dietary factors other than vitamin D play a causal role in MS, but neither can such a role be excluded, particularly for diet during adolescence or childhood, which may be important periods in the etiology of MS.

Estrogen has been hypothesized to protect against MS, because in high levels it appears to promote the non-inflammatory type 2 immune response, rather than the pro-inflammatory type 1 response predominately seen in MS, and because during pregnancy, when estrogen levels are high, women with MS experience fewer relapses than during the puerperium. 161 In prospective studies, 126, 127, 162, 163 neither oral contraceptive use, parity, nor age at first birth 162 was associated with MS risk. A decreased risk of MS during pregnancy followed by an increased risk during the first 6 months after delivery was shown in a study based on a general practice database in the United Kingdom. 163 In the same study, recent use of oral contraceptives was also associated with a reduced risk. 163 Collectively, these studies suggest that short-term exposure to estrogen may be protective against MS, but that this protection is transient.

Concerns that the hepatitis B vaccine may increase the risk of MS were raised after widespread administration of the vaccine in France, 164 but the results of most studies have not supported a causal association. Studies in the United States conducted among subjects included in a health care database, 165 among nurses, 166 and among participants in three health maintenance organizations 167 found no association between hepatitis B vaccination and risk of MS. Further, in studies of children and adolescents, no association was found between hepatitis B vaccination and MS risk 168 or risk of conversion to MS among children with a first demyelinating event. 170 However, a case-control study conducted in the General Practice Research Database in the United Kingdom did find a threefold increased risk associated with receipt of the vaccine within 3 years before MS onset, 170 and a French case-control study reported a nonsignificant increased risk of MS among individuals with clinically isolated syndrome after vaccination. 171 Among individuals with MS, the vaccine does not appear to increase the risk of relapses. 172 Overall, there is no convincing evidence that hepatitis B vaccination increases MS risk.

Other environmental factors have been associated with MS, but the available evidence is sparse, and the relevance of these factors to MS etiology remains uncertain. An increased risk of MS has been reported in relation to exposure to organic solvents, [173] [174] [175] [176] [177] physical trauma, 178 and psychological stress from the loss of a child (bereavement), 179 whereas a decreased risk has been observed for use of penicillin 180 and antihistamines, 181 high levels of serum uric acid, [182] [183] [184] and tetanus toxoid vaccination. 185 

Epidemiology: Principles and methods

The contribution of changes in the prevalence of prone sleeping position to the decline in sudden infant death syndrome in Tasmania

Geographic variation of MS incidence in two prospective studies of US women

Multiple sclerosis in US veterans of the Vietnam era and later military service: Race, sex, and geography

Sex ratio of multiple sclerosis in Canada: A longitudinal study

The Danish Multiple Sclerosis Registry: A 50-year follow-up

The epidemiology of multiple sclerosis in Europe

Epidemiology of multiple sclerosis in U.S. veterans: I. Race, sex, and geographic distribution

Epidemiology of multiple sclerosis: Incidence and prevalence rates in Denmark 1948-64 based on the Danish Multiple Sclerosis Registry

The frequency and geographic distribution of multiple sclerosis as indicated by mortality statistics and morbidity surveys in the United States and Canada

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

Latitude, migration, and the prevalence of multiple sclerosis

The incidence and prevalence of reported multiple sclerosis

Multiple sclerosis in Australia and New Zealand: Are the determinants genetic or environmental?

MS epidemiology world wide: One view of current status

Multiple sclerosis in the Japanese population

Epidemiologic evidence for multiple sclerosis as an infection

The distribution of multiple sclerosis

Clinical and epidemiological profile of multiple sclerosis in a reference center in the State of Bahia, Brazil

Multiple sclerosis in Latin America

Multiple sclerosis in KwaZulu Natal, South Africa: An epidemiological and clinical study

Prevalence of multiple sclerosis in 19 Texas counties

Incidence and prevalence of multiple sclerosis in Olmsted County

Multiple sclerosis in Isfahan, Iran

Multiple sclerosis

The dissemination of multiple sclerosis: A Viking saga? A historical essay

Migrant studies in multiple sclerosis

The age-range of risk of developing multiple sclerosis: Evidence from a migrant population in Australia

Multiple sclerosis and age at migration

Epidemiology of multiple sclerosis in US veterans: III. Migration and the risk of MS

Multiple sclerosis among immigrants in Greater London

Motor neurone disease and multiple sclerosis among immigrants to Britain

Motor neuron disease and multiple sclerosis among immigrants to England from the Indian subcontinent, the Caribbean, and East and West Africa

Multiple sclerosis among the United Kingdom-born children of immigrants from the West Indies

Multiple sclerosis among United Kingdom-born children of immigrants from the Indian subcontinent, Africa and the West Indies

Role of return migration in the emergence of multiple sclerosis in the French West Indies

Incidence and prevalence of multiple sclerosis in Olmsted County

Multiple sclerosis prevalence among Sardinians: Further evidence against the latitude gradient theory

Multiple sclerosis in Malta in 1999: An update

HLA-DRB1 and multiple sclerosis in Malta

The possible viral etiology of multiple sclerosis

Genetics of multiple sclerosis

Analysis of the 'epidemic' of multiple sclerosis in the Faroe Islands

Multiple sclerosis in the Faroe Islands: I. Clinical and epidemiological features

Multiple sclerosis in the Faroe Islands: An epitome

Multiple sclerosis and poliomyelitis

Epidemiological study of multiple sclerosis in Israel: II. Multiple sclerosis and level of sanitation

Multiple sclerosis in Australia: Socioeconomic factors

Epidemiology of multiple sclerosis in US veterans: VII. Risk factors for MS

Part III: Selected reviews. Common childhood and adolescent infections and multiple sclerosis

The effect of infections on susceptibility to autoimmune and allergic diseases

The hygiene hypothesis and multiple sclerosis

Association between parasite infection and immune responses in multiple sclerosis

Infectious mononucleosis and risk for multiple sclerosis: A meta-analysis

Multiple sclerosis and Epstein-Barr virus

Epstein-Barr virus

Epstein-Barr virus

The prevalence and clinical characteristics of MS in northern Japan

Environmental risk factors for multiple sclerosis: Part I. The role of infection

Epstein-Barr virus in pediatric multiple sclerosis

High seroprevalence of Epstein-Barr virus in children with multiple sclerosis

Epstein-barr virus antibodies and risk of multiple sclerosis: A prospective study

An altered immune response to Epstein-Barr virus in multiple sclerosis: A prospective study

Temporal relationship between elevation of Epstein Barr virus antibody titers and initial onset of neurological symptoms in multiple sclerosis

Epstein-Barr virus and multiple sclerosis: Evidence of association from a prospective study with long-term follow-up

Association between clinical disease activity and Epstein-Barr virus reactivation in MS

Absence of Epstein-Barr virus RNA in multiple sclerosis as assessed by in situ hybridisation

Is Epstein-Barr virus present in the CNS of patients with MS?

Increased frequency and broadened specificity of latent EBV nuclear antigen-1-specific T cells in multiple sclerosis

Identification of Epstein-Barr virus proteins as putative targets of the immune response in multiple sclerosis

An Epstein-Barr virus-associated superantigen

EBV-induced expression and HLA-DRrestricted presentation by human B cells of alpha B-crystallin, a candidate autoantigen in multiple sclerosis

Infection of autoreactive B lymphocytes with EBV, causing chronic autoimmune diseases

Dysregulated Epstein-Barr virus infection in the multiple sclerosis brain

Acyclovir treatment of relapsing-remitting multiple sclerosis: A randomized, placebo-controlled, double-blind study

A randomized, double-blind, placebo-controlled MRI study of anti-herpes virus therapy in MS

A randomized clinical trial of valacyclovir in multiple sclerosis

Integrating risk factors: HLA-DRB1*1501 and Epstein-Barr virus in multiple sclerosis

A single subtype of Epstein-Barr virus in members of multiple sclerosis clusters

Epstein-Barr virus genotypes in multiple sclerosis

Multiple sclerosis associated with Chlamydia pneumoniae infection of the CNS

Intrathecal antibody production against Chlamydia pneumoniae in multiple sclerosis is part of a polyspecific immune response

Infection with Chlamydia pneumoniae and risk of multiple sclerosis

Ioannidis A: Chlamydia pneumoniae infection and the risk of multiple sclerosis: A meta-analysis

Human herpesvirus 6 and multiple sclerosis: Systemic active infections in patients with early disease

Intrathecal antibody (IgG) production against human herpesvirus type 6 occurs in about 20% of multiple sclerosis patients and might be linked to a polyspecific B-cell response

Human herpesvirus 6 and multiple sclerosis: A one-year follow-up study

A putative new retrovirus associated with multiple sclerosis and the possible involvement of the Epstein-Barr virus in this disease

The Danish Multiple Sclerosis Registry: History, data collection and validity

Human coronavirus OC43 infection induces chronic encephalitis leading to disabilities in BALB/C mice

Some comments on the relationship of the distribution of multiple sclerosis to latitude, solar radiation, and other variables

The prevalence of multiple sclerosis in Australia

Geographical considerations in multiple sclerosis

Regional variation in multiple sclerosis prevalence in Australia and its association with ambient ultraviolet radiation

Sunlight and vitamin D for bone health and prevention of autoimmune diseases, cancers, and cardiovascular disease

Influence of season and latitude on the cutaneous synthesis of vitamin D3: Exposure to winter sunlight in Boston and Edmonton will not promote vitamin D3 synthesis in human skin

Safety and efficacy of increasing wintertime vitamin D and calcium intake by milk fortification

Serum 25-hydroxyvitamin D concentrations of New Zealanders aged 15 years and older

Multiple sclerosis: Vitamin D and calcium as environmental determinants of prevalence: A viewpoint. Part 1: Sunlight, dietary factors and epidemiology

The immonological functions of the vitamin D endocrine system

1,25-Dihydroxyvitamin D3 prevents the in vivo induction of murine experimental autoimmune encephalomyelitis

1,25-Dihydroxyvitamin D3 reversibly blocks the progression of relapsing encephalomyelitis: A model of multiple sclerosis

Treatment of experimental autoimmune encephalomyelitis in rat by 1,25-dihydroxyvitamin D(3) leads to early effects within the central nervous system

Mortality from multiple sclerosis and exposure to residential and occupational solar radiation: A case-control study based on death certificates

Skin cancer in people with multiple sclerosis: A record linkage study

Epidemiologic study of multiple sclerosis in Israel: I. An overall review of methods and findings

Epidemiological study of multiple sclerosis in Western Poland

Past exposure to sun, skin phenotype and risk of multiple sclerosis: A case-control study

Outdoor activities and diet in childhood and adolescence relate to MS risk above the Arctic Circle

Childhood sun exposure influences risk of multiple sclerosis in monozygotic twins

Vitamin D intake and incidence of multiple sclerosis

The use of a self-administered questionnaire to assess diet four years in the past

Food-based validation of a dietary questionnaire: The effects of week-to-week variation in food consumption

Serum 25-hydroxyvitamin D levels and risk of multiple sclerosis

The age-range of risk of developing multiple sclerosis: Evidence from a migrant population in Australia

Timing of birth and risk of multiple sclerosis: Population based study

Circulating 25-hydroxyvitamin D levels indicative of vitamin D sufficiency: Implications for establishing a new effective dietary intake recommendation for vitamin D

Estimates of optimal vitamin D status

Vitamin D supplementation, 25-hydroxyvitamin D concentrations, and safety

Human serum 25-hydroxycholecalciferol response to extended oral dosing with cholecalciferol

Epidemiologic study of multiple sclerosis in Israel

A case-control study of the association between socio-demographic, lifestyle and medical history factors and multiple sclerosis

How multiple sclerosis is related to animal illness, stress and diabetes

Environmental risk factors and multiple sclerosis: A community-based, case-control study in the province of Ferrara

Smoking is a risk factor for multiple sclerosis

Oral contraceptives and reproductive factors in multiple sclerosis incidence

The influence of oral contraceptives on the risk of mulitple sclerosis

Cigarette smoking and incidence of multiple sclerosis

Cigarette smoking and the progression of multiple sclerosis

Parental smoking at home and the risk of childhoodonset multiple sclerosis in children

Cigarette smoking and progression in multiple sclerosis

Neuropathological changes in chronic cyanide intoxication

Immunomodulatory effects of cigarette smoke

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

The epidemiology of acute respiratory infections in children and adults: A global perspective

Oral contraceptives, cigarette smoking and other factors in relation to arthritis

Reproductive factors, smoking, and the risk for rheumatoid arthritis

Smoking, obesity, alcohol consumption, and the risk of rheumatoid arthritis

Cigarette smoking increases the risk of rheumatoid arthritis: Results from a nationwide study of disease-discordant twins

Smoking and risk of rheumatoid arthritis

Smoking history, alcohol consumption, and systemic lupus erythematosus: A case-control study

Multiple sclerosis and nutrition

Diet and the geographical distribution of multiple sclerosis

Nutrition, latitude, and multiple sclerosis mortality: An ecologic study

The risk of multiple sclerosis in the U.S.A. in relation to sociogeographic features: A factor-analytic study

Correlation between milk and dairy product consumption and multiple sclerosis prevalence: A worldwide study

Nutritional factors in the aetiology of multiple sclerosis: A case-control study in Montreal, Canada

Studies on multiple sclerosis in Winnipeg, Manitoba, and New Orleans, Louisiana: II. A controlled investigation of factors in the life history of the Winnipeg patients

Epidemiological study of multiple sclerosis in western Poland

Milk consumption and multiple sclerosis-an etiological hypothesis

Risk factors in multiple sclerosis: A population-based case-control study in Hautes-Pyrenees

Nutritional Epidemiology

Dietary fat in relation to risk of multiple sclerosis among two large cohorts of women

Intakes of carotenoids, vitamin C, and vitamin E and MS risk among two large cohorts of women

A double-blind controlled trial of long chain n-3 polyunsaturated fatty acids in the treatment of multiple sclerosis

Low fat dietary intervention with omega-3 fatty acid supplementation in multiple sclerosis patients

Double-blind trial of linoleate supplementation of the diet in multiple sclerosis

Polyunsaturated fatty acids in treatment of acute remitting multiple sclerosis

Linoleic acid in multiple sclerosis: Failure to show any therapeutic benefit

Linoleic acid and multiple sclerosis: A reanalysis of three double-blind trials

Rate of pregnancy-related relapse in multiple sclerosis: Pregnancy in multiple sclerosis group

Oral contraceptives and the incidence of multiple sclerosis

Recent use of oral contraceptives and the risk of multiple sclerosis

A shadow falls on hepatitis B vaccination effort

No increase in demyelinating diseases after hepatitis B vaccination

Hepatitis B vaccination and the risk of multiple sclerosis

Vaccinations and risk of central nervous system demyelinating diseases in adults

School-based hepatitis B vaccination programme and adolescent multiple sclerosis

Hepatitis B vaccine and risk of relapse after a first childhood episode of CNS inflammatory demyelination

Recombinant hepatitis B vaccine and the risk of multiple sclerosis: A prospective study

Hepatitis B vaccination and first central nervous system demyelinating event: A case-control study

Vaccinations and the risk of relapse in multiple sclerosis. Vaccines in Multiple Sclerosis Study Group

Organic solvents and multiple sclerosis: A synthesis of the current evidence

Exposure to organic solvents and multiple sclerosis

Multiple sclerosis and organic solvents

Organic solvents and the risk of multiple sclerosis

The risk for multiple sclerosis in female nurse anaesthetists: A register based study

The relationship of MS to physical trauma and psychological stress: Report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology

The risk of multiple sclerosis in bereaved parents: A nationwide cohort study in Denmark

Antibiotic use and risk of multiple sclerosis

Allergy, histamine 1 receptor blockers, and the risk of multiple sclerosis

Uric acid levels in sera from patients with multiple sclerosis

Serum uric acid and multiple sclerosis

Serum uric acid levels of patients with multiple sclerosis and other neurological diseases

Tetanus vaccination and risk of multiple sclerosis: A systematic review

Epstein-Barr virus antibodies in multiple sclerosis

Epstein-Barr virus infection and antibody synthesis in patients with multiple sclerosis

Epstein-Barr nuclear antigen and viral capsid antigen antibody titers in multiple sclerosis

Increased prevalence and titer of Epstein-Barr virus antibodies in patients with multiple sclerosis

Viral antibody titers: Comparison in patients with multiple sclerosis and rheumatoid arthritis

The Italian Cooperative Multiple Sclerosis casecontrol study: Preliminary results on viral antibodies

The implications of Epstein-Barr virus in multiple sclerosis: A review

Altered antibody pattern to Epstein-Barr virus but not to other herpesviruses in multiple sclerosis: A population based case-control study from western Norway

Altered prevalence and reactivity of anti-Epstein-Barr virus antibodies in patients with multiple sclerosis

A role of late Epstein-Barr virus infection in multiple sclerosis

Exposure to infant siblings during early life and risk of multiple sclerosis