Epidemiology of multiple sclerosis

revue neurologique 172 (2016) 3–13

Available online at

ScienceDirect www.sciencedirect.com

Neuroepidemiology

Epidemiology of multiple sclerosis E. Leray a, T. Moreau b,*, A. Fromont b, G. Edan c a

Biostatistics and Epidemiology Department, EHESP, avenue du Professeur-Le´on-Bernard, 35000 Rennes, France Neurology Department, EA 4184, University Hospital of Dijon, 14, rue Gaffarel, 21000 Dijon, France c Neurology Department, University Hospital of Rennes, 2, rue Henri-le-Guilloux, 35000 Rennes, France b

info article

abstract

Article history:

Multiple sclerosis (MS) is the most frequently seen demyelinating disease, with a prevalence

Received 15 July 2015

that varies considerably, from high levels in North America and Europe (> 100/100,000 inha-

Received in revised form

bitants) to low rates in Eastern Asia and sub-Saharan Africa (2/100,000 population). Know-

5 October 2015

ledge of the geographical distribution of the disease and its survival data, and a better

Accepted 8 October 2015

understanding of the natural history of the disease, have improved our understanding of the

Available online 21 December 2015

respective roles of endogenous and exogenous causes of MS. Concerning mortality, in a large

Keywords:

the first 20 years of the disease, although life expectancy was reduced by 6–7 years in MS

French cohort of 27,603 patients, there was no difference between MS patients and controls in Multiple sclerosis

patients. In 2004, the prevalence of MS in France was 94.7/100,000 population, according to

Prevalence

data from the French National Health Insurance Agency for Salaried Workers (Caisse nationale

Incidence

d’assurance maladie des travailleurs Salarie´s [CNAM-TS]), which insures 87% of the French

Natural history

population. This prevalence was higher in the North and East of France. In several countries,

Risk factors

including France, the gender ratio for MS incidence (women/men) went from 2/1 to 3/1 from

Mortality

the 1950s to the 2000s, but only for the relapsing–remitting form. As for risk factors of MS, the most pertinent environmental factors are infection with Epstein-Barr virus (EBV), especially if it arises after childhood and is symptomatic. The role of smoking in MS risk has been confirmed, but is modest. In contrast, vaccines, stress, traumatic events and allergies have not been identified as risk factors, while the involvement of vitamin D has yet to be confirmed. From a genetic point of view, the association between HLA-DRB1*15:01 and a high risk of MS has been known for decades. More recently, immunogenetic markers have been identified (IL2RA, IL7RA) and, in particular thanks to studies of genome-wide associations, more than 100 genetic variants have been reported. Most of these are involved in the immune response and often associated with other autoimmune diseases. Studies of the natural history of MS suggest it is a two-phase disease: in the first phase, inflammation is focal with flares; and in the second phase, disability progresses independently of focal inflammation. This has clear implications for therapy. Age may also be a key factor in the phenotype of the disease. In conclusion, France is a high-risk country for MS, but it only slightly reduces life expectancy. MS is a multifactorial disease and the implications of immunogenetics are major. Preventative approaches might be derived from knowledge of the risk factors and natural history of the disease (smoking, vitamin D). # 2015 Elsevier Masson SAS. All rights reserved.

* Corresponding author. E-mail address: [email protected] (T. Moreau). http://dx.doi.org/10.1016/j.neurol.2015.10.006 0035-3787/# 2015 Elsevier Masson SAS. All rights reserved.

4

1.

revue neurologique 172 (2016) 3–13

Introduction

Multiple sclerosis (MS) is the most common demyelinating disease seen in high-income countries, and has a heterogeneous prevalence worldwide: it is highest in North America (140/100,000 population) and Europe (108/100,000), and lowest in East Asia (2.2/100,000 population) and sub-Saharan Africa (2.1/100,000). The global median prevalence of MS has increased from 30/100,000 in 2008 to 33/100,000 in 2013, according to a report by the MS International Federation. In Europe in particular, a North-South prevalence gradient has been described for distribution of the disease (higher in the North, lower in the South). France is located in the middle of Western Europe between zones of high MS prevalence (such as the Scandinavian countries and United Kingdom) and areas of low MS prevalence (Italy, Greece, Spain), and also appears to reflect the European epidemiology of the disease rather well. Given this situation, MS mortality, incidence and prevalence data and their evolution over time in France are important in that they provide a pertinent view of the epidemiology. Analyzing such data is an essential first step towards taking into account the considerable knowledge and numerous hypotheses regarding the cause(s) of MS and, above all, the environmental risk factors. Also, extensive databases for large groups of MS patients provide accurate information on the natural history of MS as a two-stage disease (first the focal inflammatory process, and then the second, which is independent of focal inflammatory markers). All of these descriptive and analytical epidemiological data will lead to a better understanding of the risk factors for MS, and may even have implications for therapeutic strategies.

2.

MS mortality

To our knowledge, around 40 epidemiological studies of mortality have been conducted in patients afflicted with MS. According to one published series [1–13], 70–88% of patients are still alive 25 years after clinical onset, and the median time from onset to death ranges from 24 years to > 45 years. These differences can be explained by differences in study periods, geographical areas and methodology (such as study population, duration, statistical method used). On the other hand, all of the studies [1,2,6–13], whatever their location, period or methods, showed excess mortality in MS patients compared with the general population matched for age, gender and follow-up duration. Life expectancy with MS seems to be reduced by 6 to 14 years. A large French study (SURVIMUS), which included 27,603 MS patients, showed that, during the first 20 years of the disease, survival was closely similar to that of the general population. The excess mortality was seen after this period and led to a reduction in life expectancy of about 6 to 7 years [14]. In fact, about 50–70% of deaths could be considered MSrelated, with MS as either the main cause or a contributing one. Progressive disability leads to severe handicaps, which increase the risk of death, especially by increasing the risk of infection. The second cause of death is cardiovascular, and the proportion of these deaths correlates strongly with age

distribution and therefore differs from one series to another. Deaths due to cancer are also frequent, although the risk of cancer in MS is not consistent between studies, being sometimes higher than and sometimes similar to that of controls [15]. Suicide also needs to be specifically mentioned as a cause of death in MS. Earlier studies demonstrated an elevated suicide rate among MS patients, ranging from 1.6 to 7.5 times that of the general population [2,12,13,16,17]; however, more recent studies do not support this trend and have reported suicide rates similar to or even below the expected rate [11,18,19], suggesting that an excess risk of suicide may no longer be a reality for MS patients. In the literature, factors associated with a better vital prognosis [1–3,5,7,8,11,13] include a relapsing–remitting phenotype, MS clinical onset before 25 or 30 years of age, initial symptoms such as optic neuritis and sensory problems, a low level of disability during the first years of the disease, and a long time lag between the first and second neurological episodes. Also, in recent decades, MS patients in the developed countries have experienced an increased life expectancy and decreased mortality. At the same time, however, the incidence of MS has increased, at least in women. These two phenomena have led to an increased prevalence of MS and, therefore, an increased number of patients in need of care for this chronic disease.

3.

Risk factors of MS

The cause of MS is multifactorial: both genetic and environmental risk factors contribute to disease risk. Several factors have been assessed and are reviewed below. However, the specific causes of MS are still largely unknown and, at present, there are no well-established factors to assist disease prevention [20,21].

3.1.

Environmental factors

The following data are largely derived from an umbrella review recently published by Belbasis et al. [21]. Indeed, they have provided a rigorous and systematic assessment of published reviews and meta-analyses, representing decades of research on environmental risk factors for MS (609 articles found in their search, 20 articles considered eligible). Of the 44 factors included in their analysis, only three showed strong, consistent evidence of an association with MS with no suggestion of bias: immunoglobulin G (IgG) seropositivity to Epstein-Barr virus (EBV) nuclear antigen; infectious mononucleosis; and smoking (Table 1). The three associations were statistically significant (P < 0.001) and based on > 1000 cases; the between-study heterogeneity was not large and the 95% prediction interval excluded the null value. Indeed, the strongest known risk factor for MS is infection with EBV [22], as this led to consistent results, whatever the place, period or study design. Compared with non-infected individuals, the risk of developing MS is approximately 15 times higher among those infected with EBV in childhood and about 30 times higher among those infected with EBV in adolescence or later in life. Although the mechanisms

5

revue neurologique 172 (2016) 3–13

Table 1 – Environmental risk factors for multiple sclerosis supported by convincing evidence.

Anti-EBNA IgG seropositivity Infectious mononucleosis Smoking

Sample size

Summary effect size [95% CI]

P-value

Estimated between-study heterogeneity

95% prediction intervala

> 1000 > 1000 > 1000

4.46 [3.26–6.09] 2.17 [1.97–2.39] 1.52 [1.39–1.66]

< 0.001 < 0.001 < 0.001

Not large Not large Not large

Excluding null value Excluding null value Excluding null value

Adapted from Belbasis et al., 2015 [21]. EBNA: Epstein-Barr virus nuclear antigen; IgG: immunoglobulin G. a Evaluates uncertainty of effect expected in any new study addressing that same association.

underlying this association remain unclear, the data provide strong evidence of a causal relationship between EBV infection and MS risk. Primary EBV infection at an early age is typically asymptomatic, but primary infection during adolescence or adulthood often manifests as infectious mononucleosis, which has been associated with a two- to threefold increased risk of MS [23]. In the above-mentioned meta-analysis [21], summary effect sizes for EBV and infectious mononucleosis were about 4 and 2, respectively. The association between smoking and MS is also positive, but the effect was modest (summary effect size: 1.52 [1.39– 1.66]) in the meta-analysis by Belbasis et al. [21]. In another meta-analysis, which focused only on smoking and MS [24], 14 articles that investigated MS risk following cigarette smoking were analyzed and showed a significant association. However, further work is needed to understand the mechanism behind the association and to determine how smoking interacts with other established risk factors. As also shown in the umbrella review, many studies have been conducted on risk factors for MS, but many of them had weaknesses in either their design or analysis, and their findings were not supported by convincing evidence. In addition, the Belbasis et al. meta-analysis confirmed the absence of any association between certain factors and MS. This was the case for several vaccines [tetanus, diphtheria, influenza, bacille Calmette–Gue´rin (BCG), measles–mumps– rubella (MMR), poliomyelitis, hepatitis B virus, typhoid fever], biochemical factors (the presence of dental amalgam), previous surgery and traumatic events (tonsillectomy, adenoidectomy, traumatic injury), and the presence of allergies, eczema and chronic cerebrospinal venous insufficiency (CCSVI). This well-designed analysis thus refuted a variety of factors initially thought to be significantly associated with MS. Several studies of long duration have reported differences in MS incidence or prevalence according to geographical latitude. This trend has been partly explained by the different levels of vitamin D and exposure to the sun at different latitudes. Indeed, higher serum 25(OH)D was significantly associated with lower MS incidence. However, as substantial heterogeneity has been found between studies, Belbasis et al. concluded that further prospective studies and clinical trials are needed to increase the level of evidence regarding this factor. In an era where socioeconomic inequalities affect most diseases, results regarding socioeconomic factors in MS are not clear. Goulden et al. [25] found 13 studies that reported no evidence of such an association, and three, which reported an association between MS and low socioeconomic status (SES).

These 16 studies were mostly from more egalitarian countries, and evidence of an association between a high SES and increased MS risk is inconsistent, and begs further research. The association between MS and stress or traumatic events has also been studied. In two cohorts of female nurses (the Nurses’ Health Study, in which 121,700 women were followed continuously from 1976, and the Nurses’ Health Study II, which included 116,671 followed from 1989) [26], severe stress at home was not associated with an increased risk of MS. Also, no significantly increased risk of MS was found among those who reported severe physical abuse during childhood or adolescence, or those who had been repeatedly forced into sexual activity in childhood or adolescence. Indeed, such results do not support a major role of stress in the development of the disease. Interestingly, environmental factors that are mostly encountered during childhood appear to play a role. Thus, if childhood is the most fragile period with regard to MS onset later in life, then preventative measures should be applied early in life. For example, adopting a diet rich in vitamin D, playing outdoors and avoiding passive smoking would be extremely simple measures of primary prevention as a public health strategy. However, these hypotheses need to be confirmed by prospective evaluations, which are difficult to conduct [27].

3.2.

Genetic factors

Studies involving informative cohorts, such as twins, conjugal pairs and adoptees, suggest that familial clustering is determined by genetic factors. Nevertheless, a lack of precision and the bias inherent in estimating the risk of familial recurrence limit what can be inferred on comparing risks between relatives (segregation analysis). However, it is generally suggested that the available data are consistent with a polygenic model in which risk is defined by a single moderate-effect allele [odds ratio (OR) near 3 or 4] and several alleles of much smaller effect (OR < 1.5). Familial recurrence falls geometrically with the degree of relatedness [28,29]. The association between MS and variations in the genes encoding human leukocyte antigens (HLAs) within the major histocompatibility complex was first observed several decades ago [30,31]. Large families with a large number of affected siblings per generation have a much higher than expected rate of the main risk allele for MS, HLA-DRB1*15:01, which has an OR of about 3 [32,33]. With the advent of genome-wide association studies (GWAS), identification of associations of a single nucleotide polymorphism (SNP) with IL7R and IL2RA genes has been an important step forward [34,35].

6

revue neurologique 172 (2016) 3–13

Collectively, these analyses (of tens of thousands of cases and controls) have identified 110 variants (outside the major histocompatibility complex) associated with MS susceptibility. These variants frequently implicate genes associated with immunological processes, and are often associated with other autoimmune diseases as well [36].

3.3.

Gene–environment interactions and epigenetics

As already mentioned, the precise causes of MS are unknown, and only a few genetic and environmental factors have been identified to date, with effects that are modest. Studies of the interactions between these two kinds of factors and their combined effects will probably be helpful and lead to a better understanding of the etiology of MS. Although studies of epigenetic changes in MS have only begun in the last decade, a growing body of literature suggests that epigenetic changes may be involved in the development of MS, possibly by mediating the effects of environmental risk factors such as smoking, vitamin D deficiency and EBV infection [37].

4.

Prevalence and incidence

agricole (MSA) [44], which insures 7% of the French population (mostly farmers, agricultural workers and their families). The national prevalence of MS was estimated at 65 per 105 inhabitants [62.5–67.5] on 1 January 2003. Prevalence was higher in northeastern regions (100 per 105 population) than in southwestern regions (50 per 105). Another study was based on the largest health insurance system, the CNAM-TS, which insures 87% of the French population (private salaried employees, civil and non-civil servants, and their families) [45]. On 31 October 2004, MS prevalence was 94.7 per 105 population [94.3–95.1] and was higher in northeastern France, but no obvious gradient was found when a Bayesian approach was used. This method makes it possible to take account of heterogeneity across geographical units and spatial autocorrelations. However, the main limitation of these studies was the use of a single source of data, which may have missed patients. To overcome this drawback, a study conducted in Haute-Garonne estimated MS prevalence by matching several data sources [hospitals, the public health insurance system and the MidiPyre´ne´es (MIPSEP) network] and then applying a capturerecapture method [46]. In this case, prevalence was estimated at between 110 and 149 cases per 105 inhabitants.

4.2.

Incidence

The distribution and frequency of MS are assessed by estimates of prevalence and incidence. ‘Prevalence’ refers to the proportion of a population that has the disease at or during a specified time, while ‘incidence’ refers to the proportion of new cases arising during a specific time period. Prevalence reflects the ‘stock’ of patients, while incidence reflects the dynamic dimension of the disease. As mentioned above, assessment of the different incidences and prevalences across populations can reveal ‘spatial’ (location) differences in distribution of the disease, which are important for identifying environmental and genetic factors contributing to MS. Yet, it is difficult to compare studies when they use different methodologies. A variety of sources, including hospital and clinical records, neurologists and other physicians (ophthalmologists, functional therapists) as well as MS registries and administrative databases, are used to ascertain MS cases. In such studies, the diagnosis was established by health professionals or through a review of medical records. However, MS diagnostic criteria are not always the same, as they have changed over time.

The first studies of incidence were performed at the regional scale, and was estimated at 4.17 per 105 inhabitants per year in 1976–1978 in Brittany [47], 4.22 per 105 per year in 2003–2005 in Auvergne [48] and 5.5 per 105 in 1990–2002 in Lorraine [49]. The incidence in Brittany was reassessed recently, thanks to three medical sources of case ascertainment (neurologists, Cancer Support France and a regional MS clinic) [50], and found an incidence of 4.41 per 105 population (2000–2001) after age standardization for the European population. One study estimated MS incidence in France at the national level thanks to CNAM-TS data [51]. After age standardization for the European population, the incidence was 6.8 per 105 inhabitants [6.7–6.9] between 2001 and 2007. When the under-notification rates (11.5% and 29%) were taken into account, the incidence was estimated at between 7.6 and 8.8 per 105. It appeared to be higher in northeastern France, and lower along the Atlantic coast and in the Alps, as well as on both sides of the Rhoˆne river.

4.1.

4.3.

Prevalence

Among studies that provided standardized estimations, the prevalence of MS in Europe varied from 170.5 per 105 population in the Swedish county of Va¨rmland [38], 154.5 per 105 in Denmark [39] and 163 per 105 in Seina¨joki, a district of Finland [40], to 70.6 per 105 inhabitants in Las Palmas in the Canary Islands [41]. As for incidence, estimates varied from 7.6 per 105 population in Oppland County, Norway, during 1994–98 [42] and 11.6 per 105 in Seina¨joki during 1979–93 [43] to 4.1 per 105 during 1998–2002 in Las Palmas [41]. These results seem to suggest a north to south gradient in the northern hemisphere. In France, several studies on a national scale have provided prevalence estimates. One was based on data from the second-largest health insurance system, the Mutualite´ sociale

Geographical gradient: myth or reality?

The gradient concept has been questioned [52,53]. Recent studies have revealed increases in incidence and prevalence rates in all countries in Southern Europe, and these have contributed to attenuation of the gradient [54]. This phenomenon might be explained by better case ascertainment thanks to the use of magnetic resonance imaging (MRI), new diagnostic criteria and new treatments. Concerning the increase in prevalence, this may be due to the increase in life expectancy [7]. Besides problems of case ascertainment, the statistical methodology could also contribute to the appearance of a gradient. In recent years, standardized rates from reference populations have become more frequent. A meta-analysis that included all MS incidence and prevalence

revue neurologique 172 (2016) 3–13

studies conducted between 1980 and 1998 in countries located between the northern latitudes of 40 and 60 degrees showed that the effect of latitude on standardized prevalence had decreased. In fact, the effect disappeared when the incidence was standardized [55]. As a consequence, the gradient may only be a methodological artifact. The migration of populations may also have contributed to blurring of the gradient concept. Migration might mitigate the increased MS incidence in Northern Europe, as migrants have come almost exclusively from low-risk MS areas. Environmental considerations may also affect the presence of a latitude gradient. Several studies have revealed a negative correlation between MS risk and sunshine [56,57]. A Norwegian study showed that outdoor activities in childhood reduced the risk of MS, and this applied even north of the Arctic Circle [58]. However, patterns of exposure to the sun among Northern Europeans, who have little summer sunshine, and those in Southern Europeans, who protect themselves from the sun, tend to level out any possible effects of solar radiation. In addition, several exceptions to the gradient can be found in both the northern and southern hemispheres due to genetics. In Finland, MS prevalence is higher in the south than in the north, where the Sa´mi people live. The Sa´mi have genes that protect them against MS. In Sardinia, the prevalence of MS is unusually high despite its latitude whereas, in the southern hemisphere, the prevalence of MS in New Zealand is low, despite its geographical location. This could be explained by its population of Maori. For Koch-Henriksen and Sorensen [53], the prevalence gradient in the northern hemisphere is at the limit of significance, and there is no incidence gradient. In the southern hemisphere, the prevalence gradient has also disappeared and only a marked incidence gradient, according to a few well-documented studies that involve Australia and New Zealand, persists. In a study in New Zealand, Fawcett and Skegg [59] showed an increased incidence from north to south over the 13 degrees of latitude covered by the country. The persistence of this gradient in the southern hemisphere is difficult to explain. In Australia and New Zealand, the population is highly homogeneous, and the indigenous Maori and Aborigine populations have a very low risk of MS. Although most New Zealanders have British ancestors, their risk of MS is half that of British people living in Great Britain. This preserved latitude gradient in the southern hemisphere is an important argument in favour of a strong influence of the environment on the risk of MS. In addition, a recent study has reopened the debate as to whether or not there is an MS prevalence gradient. This was a meta-analysis that involved 321 studies from 59 countries conducted between 1923 and 2009, and concluded that there is a latitude gradient for prevalence that persists even after standardization for age. The authors explained that the differences between their results and those of Koch-Henriksen and Sorensen [53] were due to better data collection, thanks to the use of several sources, the inclusion of unpublished work in English and better analyses. In this meta-analysis by Simpson et al. [60], studies were weighted by the inverse of the variance. This allowed the authors to weight studies in which the number of MS cases was low and with a wide variance. In this meta-analysis, and unlike the work of Koch-Henriksen

7

and Sorensen, studies comprising fewer than 20 MS cases were included, thus avoiding bias. In fact, disregarding studies with few MS cases may have excluded a large number of studies conducted at lower latitudes and thus blurred the notion of a gradient. In addition, in the Simpson et al. study, the prevalence data were age standardized to take into account differences in the study age ranges between zones. Prevalence was also adjusted for time period (year), which reinforced the magnitude and significance of the association between latitude and prevalence. Yet, despite these methodological advantages, the study also had weaknesses. First, the studies that were included were conducted at different centers, which led to heterogeneous case ascertainment. Furthermore, the study suffered from selection bias due to the inclusion of series prevalences estimated from the same region and the exclusion of nonpeer-reviewed studies. There was also measurement bias because the included research did not systematically use the same diagnostic criteria for MS. Finally, for the northern hemisphere, both Simpson et al. and Koch-Henriksen and Sorensen agreed that the prevalence gradient was still present, albeit very strong in the former study [60] and of borderline significance in the latter [53]. However, in the former study, the prevalence changes per 100,000 inhabitants per degree of latitude (slope) varied from one continent or country to another. While these changes were large for North America (slope = 15.35), Australia and New Zealand (slope = 8.38), they were less marked for Central Europe and the Atlantic coast (slope = 2.82). The gradient was small for Europe. Thus, it is possible that Bayesian smoothing techniques caused it to disappear in favor of areas with low and high prevalence, as the studies included in the Simpson et al. metaanalysis did not take into account either spatial heterogeneity or autocorrelations. The presence or not of the gradient therefore depends on the methodology used. In short, the concept of areas with low and high prevalences/incidences seems to be more acceptable and more important than gradients. It would be useful to speculate on the reasons for this, and to explain those areas that are over and under the mean risk to identify MS etiologies. The environment, rather than genetics, might largely explain such geographical variations [33,61,62], especially in a country like France where the population is genetically relatively homogeneous [63].

5.

Gender ratio

Over the past 30 years, the MS gender ratio has changed due to an increase in the incidence of MS in women. Between 1960 and 2005, the gender ratio adjusted for the year of MS onset increased from 1.68 to 2.45 (P = 0.017), according to data from the Lyon MS database collected by the European Database for Multiple Sclerosis (EDMUS) system. This analysis was conducted in 4495 patients, including 3030 women (67.4%). However, the change in gender ratio over time was observed only for the relapsing form of MS and not for progressive forms (P = 0.53) [64]. In Canada, the gender ratio for MS adjusted for year of birth increased from 1.9 between 1936 and 1940 to 3.2 between 1976

8

revue neurologique 172 (2016) 3–13

and 1980 [65]. The Danish National Registry showed that the incidence among women has doubled since 1970, while remaining stable among men. The same findings were made in Lorraine, France [49], in Norway [66] and in Australia [67]. Such increases in gender ratios to favor women could be explained by improvements in access to healthcare and the fact that women consult more easily than men for even minor symptoms [68]. Another explanation could be that there are more benign cases of MS in women, which are now being identified more quickly with the use of MRI and new diagnostic criteria. However, this hypothesis was not confirmed by Orton et al. in Canada [65]. In that study, the delay between the onset of symptoms and diagnosis was the same regardless of gender. The increase in gender ratio is an indirect marker of the currently increasing incidence of MS in women [49,69–74]. This might indicate a change in their environment secondary to a change in lifestyle. Several factors have been highlighted, including later pregnancy, use of oral contraception, smoking, obesity, less physical activity and more stress [75,76]. These lifestyle changes, which may lie at the root of the increased MS incidence, were explored in a case-control study conducted in Crete between 1980 and 2008 [77]. Over three decades, the gender ratio increased from 0.9 in 1980 to 2.1 in 2008. Compared with controls, women with MS were older at the time of their first pregnancy and more likely to use contraception, smoke, drink alcohol and take vitamin supplements. Also, women with MS had had more childhood infectious diseases. Furthermore, most of the women with MS lived in urban areas or had left rural areas for longer than the controls, and were more likely than controls to drink pasteurized cow’s milk and not goat’s milk. One protein in pasteurized cow’s milk, butyrophilin, can mimic myelin oligodendrocyte glycoprotein (MOG), a candidate autoantigen in MS. This molecular mimicry may explain these differences.

6.

The natural history of MS in France

Only studies of large groups of MS patients using extensive databases can give accurate information on the natural history of MS. Specifically, the requirements for such studies include identifying a geographically defined population with MS at the early stages, achieving complete ascertainment and a complete follow-up for an appropriate period, and applying outcome measurements that detect all possible relevant outcomes. No large published cohorts have completely satisfied these requirements. In Canada, epidemiological data from two major reference MS centers, one in Toronto [78–81] and one in Vancouver [82,83], have been published. They did not use the same methodology for collecting data and had significant differences in the assessment of disability progression [82]. In France, epidemiological data of the natural history of MS have been published from two reference MS centers, one in Lyon [84,85] and the other in Rennes [86], and also from a population-based cohort in the Lorraine region (LORSEP) [87]. The LORSEP is the only MS population-based registry in France and, given this nature, is probably more representative of MS than specialist centers, which may be subject to recruitment

Table 2 – Baseline characteristics of three French multiple sclerosis (MS) cohorts. Lyon Number of patients Gender (male/female, %) Mean age at onset (years) Initial symptoms Optic neuritis Long-tract symptoms Brain-stem symptoms Combination Initial course of MS Relapsing Progressive In relapsing MS > 1 relapse in first 2 years after onset Sequelae after first relapse Mean follow-up duration (years)

Rennes

LORSEP

1844 36/64 31.0 (10)

2054 30/70 (9.8)

2871 28/72 33 (10)

18% 52% 9% 21%

21% 51% 12% 16%

19% 38% 14% 28%

85% 15%

78% 22%

87% 13%

52.3%

53%

NA

17.5% 11 (10)

15% 13 (9)

15.8% 14 (9)

From references [86,87,89]. NA: not applicable.

biases (for example, more severe or complicated cases). These three MS groups were followed for > 10 years, their data were recorded on standardized computerized forms designed for EDMUS software [88] and analyzed with the same methodology. All three cohorts have provided important insights into the natural history of MS in France (see below).

6.1. Characteristics of the MS population in Lyon and Rennes Regarding the Lyon cohort [84,85,89], this computerized surveillance system was established in 1976, and included all patients with a diagnosis of MS examined at least once at the Neurological Clinic in Lyon, the regional referral center for MS for the Rhoˆne-Alpes region. Since 1990, the data have been included in EDMUS [88]. The demographic and clinical characteristics, according to initial MS course, of 1844 MS patients are presented in Table 2. During follow-up, 1026 patients (56%) reached irreversible Disability Status Scale (DSS) 4, of whom 591 (57.6%) went on to reach irreversible DSS 6. Regarding the Rennes cohort [86], patients were identified through the Rennes MS Clinic (MSC), the regional referral center for MS in Western France [8]. Patients referred to the MSC mainly live in Brittany (60%), and the Pays de Loire region (20%) and bordering regions (14%), with the remainder coming from other French or European regions (6%). Since January 1976, any new case of MS referred to the Rennes MSC has been systematically registered, whatever the date of the first MS symptoms. In 1996, these data were included in EDMUS [88]. The demographic and clinical characteristics, according to initial MS course, of 2054 patients are also presented in Table 2. During the followup period, 1415 patients (68.9%) reached irreversible DSS 3, of whom 718 patients (51.7%) went on to irreversible DSS 6. During follow-up, 618 patients with relapsing-onset MS (38.4%) converted to the secondary progressive phase after a median time from MS clinical onset of 16.0 years [14.7–17.3] and at a median age of 40.4 years [39.3–41.4]. These two cohorts, however, do not reflect the pure natural history of MS, as some patients received disease-modifying

9

revue neurologique 172 (2016) 3–13

therapies. Since the late 1960s, azathioprine and cyclophosphamide have been used to treat MS in France. Azathioprine is administered mainly in relapsing–remitting MS (RRMS) and after the third relapse, and is then usually stopped when the disease becomes progressive. Cyclophosphamide therapy is used only in severe cases or during the progressive phase of the disease. In Lyon, a total of 903 patients (49%) had received one or more drugs: azathioprine (n = 820 patients); cyclophosphamide (n = 78 patients); methotrexate (n = 60); interferon beta (n = 72); methotrexate (n = 60); and mitoxantrone (n = 18). Treatments were given for a short period of time relative to the overall duration of the disease in any given patient, and treatment responses did not significantly affect the results of analyses of disability progression [89]. In Rennes, a total of 1154 patients (56.2%) received diseasemodifying drugs (DMDs) for at least 6 months, accounting for 4515 patient-years (17.2% of the total number of patient-years; 16.9% with relapsing-onset MS; and 18.1% with progressiveonset MS; non-significant difference). Treatment was started on average 7.7  7.1 years after MS clinical onset, and included interferon beta (28.5%), mitoxantrone (26.7%), azathioprine (21.4%), methotrexate (13.0%), cyclophosphamide (6.7%) and glatiramer acetate (2.2%). Interestingly, on comparing a subgroup of partially treated patients with a subgroup of never-treated patients, no significant differences were observed for disability progression in either relapsing-onset or progressive-onset MS patients. In the LORSEP registry [87], patients were identified through the Lorraine MS regional network, which comprises neurologists, MS centers, radiologists, biologists, nurses, physiotherapists and the MS Association, making the recruitment population-based. Data were entered into the EDMUS system. The mean duration of follow-up was nearly 14 years; during the follow-up of 2871 patients, 1795 (63%), 1416 (49%) and 817 (28%) patients reached DSS scores of 3, 4 and 6, respectively. Among the 2518 patients with relapsing-onset MS, 933 (37%) had a secondary progressive course at the time of analysis. Some LORSEP patients had also received DMDs, but no details were provided.

6.2. Disability progression in Lyon, Rennes and LORSEP MS cohorts As briefly presented above and in Table 3, the proportions of patients who reached disability milestones, and the time to reach such scores, were closely similar across the three cohorts. Median times to reach DSS 4 and DSS 6 from clinical onset were slightly longer in the LORSEP cohort and were probably linked to the population-based recruitment. The LORSEP team also showed that patients followed at an MS center had earlier disability than outpatients or patients managed otherwise [90].

6.3.

Two-stage disability progression in MS

Hypothesizing that the occurrence of disability or early irreversible disability might correspond to a key step in the disease process, Leray et al. [86] defined two phases in the course of MS: an early ‘phase 1’, from clinical onset to irreversible DSS 3; and a later ‘phase 2’, from irreversible DSS 3 to irreversible DSS 6. Also, to explore the relationship between disability progression during phase 1 vs phase 2, MS patients who had reached irreversible DSS 3 were classified into five subgroups according to phase 1 duration in years: subgroup 1 = 0 to < 3; subgroup 2 = 3 to < 6; subgroup 3 = 6 to < 10; subgroup 4 = 10 to < 15; and subgroup 5  15. These thresholds were selected to obtain groups of comparable size and to allow statistical comparisons: subgroup 1 = 523 patients; subgroup 2 = 290 patients; subgroup 3 = 254 patients; subgroup 4 = 172 patients; and subgroup 5 = 176 patients. Of the 718 patients who reached DSS 6, the mean time to reach DSS 6 from DSS 3 was 5.47 years, which was nearly identical in the five subgroups, in the whole MS population (between 4.88 years and 5.74 years, P = 0.764), in relapsing-onset MS (between 4.93 years and 6.31 years, P = 0.394) and in progressive-onset MS (between 4.74 years and 6.10 years, P = 0.444). These results indicate that disability progression during phase 2 is independent of disability progression during phase 1 (Fig. 1).

Table 3 – Kaplan–Meier estimates of median time (years) to irreversible disability scores according to initial course of multiple sclerosis (MS) in cohorts from Lyon [89], Rennes [86] and LORSEP [87]. Relapsing MS Lyon From CO to DSS 3 DSS 4 DSS 6 DSS 7 DSS 3 to DSS 6 DSS 4 to DSS 6 DSS 4 to DSS 7 Median age at DSS 3 DSS 4 DSS 6

Rennes

Progressive MS LORSEP

Lyon

15.9 24.5

0 7.1 13.4

3.0

5.4 12.0

10.0 11.4 23.1 33.1

21.7

CO: clinical onset; DSS: Disability Status Scale.

51.1

10.0

3.0 10.3

6.4

42.4 44.8 55.3

LORSEP

2.0

7.4 5.7 12.1

Rennes

4.0

41.9 42.1 53.0

52.1

10

revue neurologique 172 (2016) 3–13

only relapsing-onset MS, and were younger at DSS 3 with both phenotypes. In contrast, gender had no influence on the duration of phase 2, whatever the disease phenotype at onset.

6.7.

A strong correlation was found between age at onset and disability progression in the whole MS population, especially with relapsing-onset MS. While patients who were younger at onset were also younger when they reached DSS 3 and 6, they also exhibited slower disability progression during phase 1. In contrast, age at onset did not influence the duration of phase 2.

Fig. 1 – Disability progression during phase 2 (mean time from DSS 3 to DSS 6) in five subgroups, defined according to duration of phase 1 (mean time from multiple sclerosis clinical onset to DSS 3), in the Rennes European Database for Multiple Sclerosis (EDMUS) cohort [86]. DSS: Disability Status Scale.

6.4.

6.8.

Predictive factors of disability progression

7.

Lessons from the natural history of MS

Understanding the dissociation between early and late disability progression, and the role of relapses, is central to the debate on the putative mechanisms of disability progression in MS. Some studies [79,81,84] have suggested that age is a key factor (if not the only one) in the natural history of MS, leading to the concept of MS as a single-stage disorder with chronic age-related neurodegeneration following onset of the disease [85]. However, the Rennes cohort [86] suggests a two-stage disorder, with a first stage during which focal inflammatory lesions (with relapse as a clinical marker) influence disability progression, and a second stage during which disability progression seems independent of focal inflammatory markers. In contrast, with the progressive-onset phenotype, focal inflammatory lesions are clinically asymptomatic for a long period of time and only detectable on MRI, thereby restricting clear identification of the expected first stage of the disease. The concept of MS as a two-stage disease is also supported by MRI data [91], especially the plateau relationship between

Phenotype

Disease phenotype at MS onset was found to correlate not only with age at onset, but also with disability progression. Although patients with relapsing and progressive-onset roughly reached DSS 3 and DSS 6 at the same age, the data showed that the durations of phase 1 and, to a lesser extent, of phase 2 were significantly shorter with the progressive-onset phenotype than the relapsing-onset phenotype.

6.6.

Relapse history in relapsing-onset MS

In relapsing-onset MS, the presence of a residual deficit after the first relapse and the occurrence of relapses during the first 2 or 5 years of MS significantly shortened the duration of phase 1, but did not influence disability progression during phase 2. During phase 2, disability progression was more influenced by previous conversion to secondary progression than by the occurrence of relapses.

In the three French cohorts, their initial characteristics (gender, age, residual deficit following first relapse, and number of relapses during the first 2 or 5 years of MS) were assessed as potential prognostic factors during early disease (between clinical onset and DSS 3 or 4) as well as during later stages of disease (between DSS 3, 4 and 6). The results were almost identical (Table 4) and showed that, while most of these factors were predictive of the duration of phase 1, most of them also had no influence on the duration of phase 2 of the disease.

6.5.

Age at clinical onset

Gender

Men were younger than women at clinical onset of only progressive-onset MS and had a shorter phase 1 duration for

Table 4 – Significant (S) and non-significant (NS) predictive factors of disability progression during phase 1 and phase 2 in the Lyon, Rennes and LORSEP cohorts from EDMUS. Phase 1 (CO to DSS 3/4) Factors Gender Age at onset Onset MS phenotype Initial MS symptoms Sequelae from first relapse Number of relapses for first 2 or 5 years of MS

Lyon S S S S S S

Rennes *

NS S S S S S

Phase 2 (DSS 3/4 to DSS 6)

LORSEP

Lyon

Rennes

LORSEP

S S NA S S S

NS NS NS NS NS NS

NS NS S NS S NS

NS S NA NS NS NS

EDMUS: European Database for Multiple Sclerosis; CO: clinical onset; DSS: Disability Status Scale; NA: not available. P = 0.06 for phase 1 duration (close to statistical significance).

*

revue neurologique 172 (2016) 3–13

T2 burden of the disease and disability according to an expanded DSS value > 4.5, and the strong correlation between changes in T2 lesion load within the first 5 years of MS and disability status at 20 years of disease duration [92]. Indeed, the concept of MS as a two-stage disease based on these epidemiological data has clear implications for therapeutic strategies in MS, and reinforces the idea of a therapeutic window of opportunity as suggested by Coles et al. [93].

8.

Conclusion

Given the mortality, prevalence, incidence, evolution of the gender ratio and geographical distribution of MS in France, this country can now be more precisely defined as a high-risk country for MS. Improved knowledge of the natural history of MS, with the dissociation between early and late disability progression and the impact of relapses on disability progression, has led to a better understanding of MS mechanisms and better guidance for therapeutic strategies. As stated by Simpson et al. in 2015 [94], there is evidence that epidemiology will continue to play a crucial role in unraveling the architecture of MS causation and clinical course. ‘‘While classic epidemiological methods are ongoing, novel avenues for research include gene–environment interaction studies, the world of ‘-omic’ research, and the utilization of mobile and social media tools to both access and track study populations, which means that the epidemiological discoveries of the past century may be but a glimpse of our understanding in the next few decades.’’

Disclosure of interest The authors declare that they have no competing interest.

references

[1] Phadke JG. Survival pattern and cause of death in patients with multiple sclerosis: results from an epidemiological survey in northeast Scotland. J Neurol Neurosurg Psychiatry 1987;50(5):523–31. [2] Sadovnick AD, Ebers GC, Wilson RW, Paty DW. Life expectancy in patients attending multiple sclerosis clinics. Neurology 1992;42(5):991–4. [3] Runmarker B, Andersen O. Prognostic factors in a multiple sclerosis incidence cohort with twenty-five years of followup. Brain 1993;116(Pt 1):117–34. [4] Kantarci O, Siva A, Eraksoy M, Karabudak R, Sutlas N, Agaoglu J, et al. Survival and predictors of disability in Turkish MS patients. Turkish Multiple Sclerosis Study Group (TUMSSG). Neurology 1998;51(3):765–72. [5] Cottrell DA, Kremenchutzky M, Rice GP, Koopman WJ, Hader W, Baskerville J, et al. The natural history of multiple sclerosis: a geographically based study 5. The clinical features and natural history of primary progressive multiple sclerosis. Brain 1999;122(Pt 4):625–39. [6] 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(Pt 8):1677–87.

11

[7] Bronnum-Hansen H, Koch-Henriksen N, Stenager E. Trends in survival and cause of death in Danish patients with multiple sclerosis. Brain 2004;127(Pt 4):844–50. [8] Leray E, Morrissey S, Yaouanq J, Coustans M, Le Page E, Chaperon J, et al. Long-term survival of patients with multiple sclerosis in West France. Mult Scler 2007;13(7):865– 74. [9] Grytten Torkildsen N, Lie SA, Aarseth JH, Nyland H, Myhr KM. Survival and cause of death in multiple sclerosis: results from a 50-year follow-up in Western Norway. Mult Scler 2008;14(9):1191–8. [10] Smestad C, Sandvik L, Celius EG. Excess mortality and cause of death in a cohort of Norwegian multiple sclerosis patients. Mult Scler 2009;15(11):1263–70. [11] Sumelahti ML, Hakama M, Elovaara I, Pukkala E. Causes of death among patients with multiple sclerosis. Mult Scler 2010;16(12):1437–42. [12] Lalmohamed A, Bazelier MT, Van Staa TP, Uitdehaag BM, Leufkens HG, De Boer A, et al. Causes of death in patients with multiple sclerosis and matched referent subjects: a population-based cohort study. Eur J Neurol 2012;19(7):1007–14. [13] Kingwell E, van der Kop M, Zhao Y, Shirani A, Zhu F, Oger J, et al. Relative mortality and survival in multiple sclerosis: findings from British Columbia, Canada. J Neurol Neurosurg Psychiatry 2011;83(1):61–6. [14] Leray E, Vukusic S, Debouverie M, Clanet M, Brochet B, Seze J, et al. Excess mortality in patients with multiple sclerosis starts at 20 years from clinical onset: data from a largescale French observational study. PLoS One 2015;10(7):e0132033. [15] Marrie RA, Reider N, Cohen J, Stuve O, Trojano M, Sorensen PS, et al. A systematic review of the incidence and prevalence of cancer in multiple sclerosis. Mult Scler 2015;21(3):294–304. [16] Ragonese P, Aridon P, Salemi G, D’Amelio M, Savettieri G. Mortality in multiple sclerosis: a review. Eur J Neurol 2008;15(2):123–7. [17] Scalfari A, Knappertz V, Cutter G, Goodin DS, Ashton R, Ebers GC. Mortality in patients with multiple sclerosis. Neurology 2013;81(2):184–92. [18] Redelings MD, McCoy L, Sorvillo F. Multiple sclerosis mortality and patterns of comorbidity in the United States from 1990 to 2001. Neuroepidemiology 2006;26(2):102–7. [19] Hirst C, Swingler R, Compston DA, Ben-Shlomo Y, Robertson NP. Survival and cause of death in multiple sclerosis: a prospective population-based study. J Neurol Neurosurg Psychiatry 2008;79(9):1016–21. [20] Ramagopalan SV, Dobson R, Meier UC, Giovannoni G. Multiple sclerosis: risk factors, prodromes, and potential causal pathways. Lancet Neurol 2010;9(7):727–39. [21] Belbasis L, Bellou V, Evangelou E, Ioannidis JP, Tzoulaki I. Environmental risk factors and multiple sclerosis: an umbrella review of systematic reviews and meta-analyses. Lancet Neurol 2015;14(3):263–73. [22] Ascherio A. Environmental factors in multiple sclerosis. Expert Rev Neurother 2013;13(12 Suppl.):3–9. [23] Ascherio A, Munger KL. Epstein-barr virus infection and multiple sclerosis: a review. J Neuroimmune Pharmacol 2010;5(3):271–7. [24] Handel AE, Williamson AJ, Disanto G, Dobson R, Giovannoni G, Ramagopalan SV. Smoking and multiple sclerosis: an updated meta-analysis. PLoS One 2011;6(1):e16149. [25] Goulden R, Ibrahim T, Wolfson C. Is high socioeconomic status a risk factor for multiple sclerosis? A systematic review. Eur J Neurol 2014;22(6):899–911.

12

revue neurologique 172 (2016) 3–13

[26] Riise T, Mohr DC, Munger KL, Rich-Edwards JW, Kawachi I, Ascherio A. Stress and the risk of multiple sclerosis. Neurology 2011;76(22):1866–71. [27] Pantazou V, Schluep M, Du Pasquier R. Environmental factors in multiple sclerosis. Presse Med 2015;44(4 Pt 2):e113–20. [28] Willer CJ, Dyment DA, Risch NJ, Sadovnick AD, Ebers GC. Twin concordance and sibling recurrence rates in multiple sclerosis. Proc Natl Acad Sci U S A 2003;100(22):12877–82. [29] O’Gorman C, Lin R, Stankovich J, Broadley SA. Modelling genetic susceptibility to multiple sclerosis with family data. Neuroepidemiology 2012;40(1):1–12. [30] Naito S, Namerow N, Mickey MR, Terasaki PI. Multiple sclerosis: association with HL-A3. Tissue Antigens 1972;2(1):1–4. [31] Jersild C, Dupont B, Fog T, Hansen GS, Nielsen LS, Thomsen M, et al. Histocompatibility-linked immune-response determinants in multiple sclerosis. Transplant Proc 1973;5(4):1791–6. [32] Olerup O, Hillert J. HLA class II-associated genetic susceptibility in multiple sclerosis: a critical evaluation. Tissue Antigens 1991;38(1):1–15. [33] Sawcer S. The major cause of multiple sclerosis is environmental: genetics has a minor role – no. Mult Scler 2011;17(10):1174–5. [34] Hafler DA, Compston A, Sawcer S, Lander ES, Daly MJ, De Jager PL, et al. Risk alleles for multiple sclerosis identified by a genome-wide study. N Engl J Med 2007;357(9):851–62. [35] Gregory SG, Schmidt S, Seth P, Oksenberg JR, Hart J, Prokop A, et al. Interleukin 7 receptor alpha chain (IL7R) shows allelic and functional association with multiple sclerosis. Nat Genet 2007;39(9):1083–91. [36] Sawcer S, Franklin RJ, Ban M. Multiple sclerosis genetics. Lancet Neurol 2014;13(7):700–9. [37] Koch MW, Metz LM, Kovalchuk O. Epigenetic changes in patients with multiple sclerosis. Nat Rev Neurol 2012;9(1):35–43. [38] Bostrom I, Callander M, Kurtzke JF, Landtblom AM. High prevalence of multiple sclerosis in the Swedish county of Varmland. Mult Scler 2009;15(11):1253–62. [39] Bentzen J, Flachs EM, Stenager E, Bronnum-Hansen H, Koch-Henriksen N. Prevalence of multiple sclerosis in Denmark 1950–2005. Mult Scler 2010;16(5):520–5. [40] Sarasoja T, Wikstrom J, Paltamaa J, Hakama M, Sumelahti ML. Occurrence of multiple sclerosis in central Finland: a regional and temporal comparison during 30 years. Acta Neurol Scand 2004;110(5):331–6. [41] Aladro Y, Alemany M, Perez-Vieitez M, Amela R, Conde M, Reyes M, et al. Prevalence and incidence of multiple sclerosis in Las Palmas, Canary Islands, Spain. Neuroepidemiology 2005;24(1–2):70–5. [42] Risberg G, Aarseth JH, Nyland H, Lauer K, Myhr KM, Midgard R. Prevalence and incidence of multiple sclerosis in Oppland County: a cross-sectional population-based study in a landlocked county of Eastern Norway. Acta Neurol Scand 2010;124(4):250–7. [43] Sumelahti ML, Tienari PJ, Hakama M, Wikstrom J. Multiple sclerosis in Finland: incidence trends and differences in relapsing remitting and primary progressive disease courses. J Neurol Neurosurg Psychiatry 2003;74(1):25–8. [44] Vukusic S, Van Bockstael V, Gosselin S, Confavreux C. Regional variations in the prevalence of multiple sclerosis in French farmers. J Neurol Neurosurg Psychiatry 2007;78(7):707–9. [45] Fromont A, Binquet C, Sauleau E, Fournel I, Bellisario A, Adnet J, et al. Geographic variations of multiple sclerosis in France. Brain 2010;133:1889–99. [46] Sagnes-Raffy C, Gourraud P-A, Hannon V, Bourrel R, Laffontan M-A, Gaulene M-C, et al. Multiple sclerosis in

[47]

[48]

[49]

[50]

[51]

[52]

[53]

[54]

[55]

[56]

[57]

[58]

[59]

[60]

[61]

[62]

[63]

[64]

[65]

Haute-Garonne: an important underestimation of case numbers. Rev Epidemiol Sante Publique 2010;58(1):23–31. Gallou M, Madigand M, Masse L, Morel G, Oger J, Sabouraud O. [Epidemiology of multiple sclerosis in Brittany]. Presse Med 1983;12(16):995–9. Taithe F, Tilignac C, Rieu L, Murat C, Clavelou P. Incidence of Multiple Sclerosis in Auvergne (French Country): a population-based study. Mult Scler 2006;12(Suppl. 1):S132. Debouverie M, Pittion-Vouyovitch S, Louis S, Roederer T, Guillemin F. Increasing incidence of multiple sclerosis among women in Lorraine, Eastern France. Mult Scler 2007;13(8):962–7. Yaouanq J, Tron I, Kerbrat A, Leray E, Hamonic S, Merienne M, et al. Register-based incidence of multiple sclerosis in Brittany (north-western France), 2000–2001. Acta Neurol Scand 2014;131(5):321–8. Fromont A, Binquet C, Sauleau E, Fournel I, Despalins R, Rollot F, et al. National estimate of multiple sclerosis incidence in France (2001–2007). Mult Scler 2012;18:1108–15. Koch-Henriksen N, Sorensen PS. The changing demographic pattern of multiple sclerosis epidemiology. Lancet Neurol 2010;9(5):520–32. Koch-Henriksen N, Sorensen PS. Why does the north-south gradient of incidence of multiple sclerosis seem to have disappeared on the northern hemisphere? J Neurol Sci 2011;311(1–2):58–63. Fernandez O, Fernandez V, Guerrero M, Leon A, LopezMadrona J, Alonso A, et al. Multiple sclerosis prevalence in Malaga, Southern Spain estimated by the capturerecapture method. Mult Scler 2011;18:372–6. Zivadinov R, Iona L, Monti-Bragadin L, Bosco A, Jurjevic A, Taus C, et al. The use of standardized incidence and prevalence rates in epidemiological studies on multiple sclerosis. A meta-analysis study. Neuroepidemiology 2003;22(1):65–74. Beretich B, Beretich T. Explaining multiple sclerosis prevalence by ultraviolet exposure: a geospatial analysis. Int MS J 2009;15(8):891–8. Orton SM, Wald L, Confavreux C, Vukusic S, Krohn J, Ramagopalan SV, et al. Association of UV radiation with multiple sclerosis prevalence and sex ratio in France. Neurology 2011;76:425–31. Kampman MT, Wilsgaard T, Mellgren SI. Outdoor activities and diet in childhood and adolescence relate to MS risk above the Arctic Circle. J Neurol 2007;254(4):471–7. Fawcett J, Skegg DC. Geographic distribution of MS in New Zealand: evidence from hospital admissions and deaths. Neurology 1988;38(3):416–8. Simpson Jr S, Blizzard L, Otahal P, Van der Mei I, Taylor B. Latitude is significantly associated with the prevalence of multiple sclerosis: a meta-analysis. J Neurol Neurosurg Psychiatry 2011;82(10):1132–41. Hutchinson M. The major cause of multiple sclerosis is environmental: genetics has a minor role – commentary. Mult Scler 2011;17(10):1176. Taylor BV. The major cause of multiple sclerosis is environmental: genetics has a minor role–yes. Mult Scler 2011;17(10):1171–3. Olerup O, Troye-Blomberg M, Schreuder GM, Riley EM. HLADR and -DQ gene polymorphism in West Africans is twice as extensive as in north European Caucasians: evolutionary implications. Proc Natl Acad Sci U S A 1991;88(19):8480–4. Debouverie M, Vukusic S, Ritleng C, Laforest L, Van Ganse E, Confavreux C. Change of demographic data over time in Multiple Sclerosis. The Lyon multiple sclerosis cohort experience: Part I: sex ratio. Mult Scler 2008;14(Suppl. 1):P116. Orton SM, Herrera BM, Yee IM, Valdar W, Ramagopalan SV, Sadovnick AD, et al. Sex ratio of multiple sclerosis in

revue neurologique 172 (2016) 3–13

[66]

[67]

[68]

[69]

[70]

[71]

[72]

[73]

[74] [75] [76] [77]

[78]

[79]

[80]

Canada: a longitudinal study. Lancet Neurol 2006;5(11): 932–6. Midgard R, Riise T, Svanes C, Kvale G, Nyland H. Incidence of multiple sclerosis in More and Romsdal, Norway from 1950 to 1991. An age-period-cohort analysis. Brain 1996;119:203–11. Barnett MH, Williams DB, Day S, Macaskill P, McLeod JG. Progressive increase in incidence and prevalence of multiple sclerosis in Newcastle, Australia: a 35-year study. J Neurol Sci 2003;213(1–2):1–6. Owens G. Gender differences in health care expanditures, resource utilization, and quality of care. J Manag Care Pharm 2008;14(3):2–6. Hernan MA, Olek MJ, Ascherio A. Geographic variation of MS incidence in two prospective studies of US women. Neurology 1999;53(8):1711–8. Kro¨kki O, Bloigu R, Reunanen M, Remes A. Increasing incidence of multiple sclerosis in Women in Northern Finland. Int MS J 2011;17(2):133–8. McDonnell GV, Hawkins SA. High incidence and prevalence of multiple sclerosis in south east Scotland: evidence of a genetic predisposition. J Neurol Neurosurg Psychiatry 1999;66(3):411. Noonan CW, Kathman SJ, White MC. Prevalence estimates for MS in the United States and evidence of an increasing trend for women. Neurology 2002;58(1):136–8. Pugliatti M, Rosati G, Carton H, Riise T, Drulovic J, Vecsei L, et al. The epidemiology of multiple sclerosis in Europe. Eur J Neurol 2006;13(7):700–22. Wasay M, Khatri IA, Khealani B, Sheerani M. MS in Asian countries. Int MS J 2006;13(2):58–65. Azarbad L, Gonder-Frederick L. Obesity in women. Psychiatr Clin North Am 2010;33:423–40. Pi-Sunyer X. The medical risks of obesity. Postgrad Med 2009;121:21–33. Kotzamani D, Panou T, Mastorodemos V, Tzagournissakis M, Nikolakaki H, Spanaki C, et al. Rising incidence of multiple sclerosis in females associated with urbanization. Neurology 2012;78(22):1728–35. Weinshenker BG, Bass B, Rice GP, Noseworthy J, Carriere W, Baskerville J, et al. The natural history of multiple sclerosis: a geographically based study. I. Clinical course and disability. Brain 1989;112(Pt 1):133–46. Kremenchutzky M, Rice GP, Baskerville J, Wingerchuk DM, Ebers GC. The natural history of multiple sclerosis: a geographically based study 9: observations on the progressive phase of the disease. Brain 2006;129(Pt 3):584–94. Scalfari A, Neuhaus A, Degenhardt A, Rice GP, Muraro PA, Daumer M, et al. The natural history of multiple sclerosis: a

[81]

[82]

[83]

[84]

[85] [86]

[87]

[88]

[89]

[90]

[91]

[92]

[93]

[94]

13

geographically based study 10: relapses and long-term disability. Brain 2010;133(Pt 7):1914–29. Scalfari A, Neuhaus A, Daumer M, Ebers GC, Muraro PA. Age and disability accumulation in multiple sclerosis. Neurology 2011;77(13):1246–52. Tremlett H, Paty D, Devonshire V. Disability progression in multiple sclerosis is slower than previously reported. Neurology 2006;66(2):172–7. Tremlett H, Yousefi M, Devonshire V, Rieckmann P, Zhao Y. Impact of multiple sclerosis relapses on progression diminishes with time. Neurology 2009;73(20):1616–23. Confavreux C, Vukusic S, Moreau T, Adeleine P. Relapses and progression of disability in multiple sclerosis. N Engl J Med 2000;343(20):1430–8. Confavreux C, Vukusic S. Natural history of multiple sclerosis: a unifying concept. Brain 2006;129(Pt 3):606–16. Leray E, Yaouanq J, Le Page E, Coustans M, Laplaud D, Oger J, et al. Evidence for a two-stage disability progression in multiple sclerosis. Brain 2010;133(Pt 7):1900–13. Debouverie M, Pittion-Vouyovitch S, Louis S, Guillemin F. Natural history of multiple sclerosis in a population-based cohort. Eur J Neurol 2008;15(9):916–21. Confavreux C, Compston DA, Hommes OR, McDonald WI, Thompson AJ. EDMUS, a European database for multiple sclerosis. J Neurol Neurosurg Psychiatry 1992;55(8):671–6. Confavreux C, Vukusic S, Adeleine P. Early clinical predictors and progression of irreversible disability in multiple sclerosis: an amnesic process. Brain 2003;126(Pt 4):770–82. Debouverie M, Laforest L, Van Ganse E, Guillemin F. Earlier disability of the patients followed in multiple sclerosis centers compared to outpatients. Mult Scler 2009;15(2): 251–7. Li DK, Held U, Petkau J, Daumer M, Barkhof F, Fazekas F, et al. MRI T2 lesion burden in multiple sclerosis: a plateauing relationship with clinical disability. Neurology 2006;66(9):1384–9. Fisniku LK, Brex PA, Altmann DR, Miszkiel KA, Benton CE, Lanyon R, et al. Disability and T2 MRI lesions: a 20-year follow-up of patients with relapse onset of multiple sclerosis. Brain 2008;131(Pt 3):808–17. Coles AJ, Cox A, Le Page E, Jones J, Trip SA, Deans J, et al. The window of therapeutic opportunity in multiple sclerosis: evidence from monoclonal antibody therapy. J Neurol 2006;253(1):98–108. Simpson Jr S, Taylor BV, van der Mei I. The role of epidemiology in MS research: past successes, current challenges and future potential. Mult Scler 2015;21(8):969– 77.