Prevalence and disease disability in immigrants with multiple sclerosis in Malm ¨ o, southern Sweden

Background: Multiple sclerosis (MS) is the most common chronic demyelinating disease of the central nervous system and the major non-traumatic cause of permanent disability in young adults. Several migration studies have been performed over the years suggesting a pattern of higher disease disability in certain ethnic groups. To our knowledge, differences in disease progression in immigrants have not been studied in Sweden before. Thus, the aims of our study were to estimate the prevalence of multiple sclerosis among first-generation immigrants in the City of Malm ¨ o and to compare differences in disease severity with the native population. Methods: All persons with multiple sclerosis living in Malm ¨ o on prevalence day 31 Dec 2010 were included. Cases were classified according to the country of birth into Scandinavians, Western and non-Western. Results: The crude prevalence was 100/100,000 (95% CI, 80 – 124) among first-generation immigrants, 154/ 100,000 (95% CI, 137 – 173) among individuals with Scandinavian background, 123/100,000 (95% CI, 94 – 162) in the Western group and 76/100,000 (95% CI, 53 – 108) in the non-Western group. The mean Multiple Sclerosis Severity Score (MSSS) value among Scandinavians was 4.2 (SD 3.5), whereas the figures in the immigrant group were 4.6 (SD 3.3) and 5.2 (SD 3.7) among Westerns respectively non-Westerns, which differences were not statistically significant. When adjusting for gender, age at onset and initial disease course, the mean MSSS difference between the non-Western and the Scandinavian individuals was 1.7 (95% CI 0.18 – 3.3, p = 0.030). There were no differences on time to diagnosis or the time from diagnosis to treatment initiation between the three groups. Conclusions: We found a lower prevalence among Western and non-Western first-generation immigrants compared to the Scandinavian population and a more severe disease in non-Western immigrants than in Scandinavians.


Introduction
Multiple sclerosis (MS) is a chronic, inflammatory, demyelinating, and neurodegenerative disease of the central nervous system.MS typically begins in early adulthood and is the most common non-traumatic cause of permanent disability in young adults [1].It is more common in women and the female to male ratio has been increasing over time being now close to 3:1 in most developed countries [2].The vast majority of patients with MS initially follows a relapsing-remitting course (RRMS), defined by acute exacerbations (relapses), with periods of relative clinical stability in between [3].In the untreated patient, secondary progressive MS (SPMS) typically develops 10-15 years after the initial event, with a gradual evolution from relapses to slowly progressive disease [2].About 5-15% of cases present as primary-progressive MS (PPMS) with gradually increasing neurological disability from onset.Individuals with PPMS are usually older at onset and the female preponderance less pronounced [4] or reversed [5].MS has serious negative effects on health-, social-, and work-related issues for the patients and their families, thus causing significant socioeconomic burden [6].
MS is a multifactorial disease caused by a complex interplay of genetic susceptibility and environmental factors.Environmental factors with the strongest evidence for involvement in MS are Epstein-Barr virus infection, tobacco exposure, lack of sun exposure, low D vitamin levels and adolescence overweight [1].
The prevalence of MS varies between countries with higher latitudes correlating with increased prevalence globally [7].Several migration studies have been performed over the years [8].Traditionally it has been claimed that migration from a high-to a low-risk area is associated with a reduction in risk, whereas migration from low-to high-risk areas is not associated with much change in risk [9].This theory has been challenged in more recent studies [10][11][12][13].Nationwide studies from high-risk countries such as Sweden and Norway have suggested that certain immigrants' groups have a lower prevalence than the native population but higher than the one in their country of origin [11,12].Likewise, a recent nationwide survey from Denmark have shown a lower MS risk among first-generation immigrants than in native Danes, but higher than in their country of origin [13].
It has been suggested that migration before the age of 15 years is critical to acquire the risk in the new country of residence both when moving from a high-to a low-risk area [14,15] and from a low-to a high-risk one [16,17].Recently, a large population-based study conducted in Ontario, Canada showed a gradually decreasing risk of MS with increasing age at migration and a rising risk with the duration of exposure to a high-risk environmental [18].
Migration studies have also studied the pattern of disease severity.North African immigrants in France are reported to have a more rapid disease progression than Europeans [19,20].Similarly, non-Western immigrants with MS in Norway are reported to have an increased disease severity compared with native Norwegians and immigrants from Western countries [21].These pattern of higher disease disability in certain ethnic groups has also been suggested in other studies [22][23][24].
To our knowledge, differences in disease progression in immigrants have not been studied in Sweden before.On the prevalence day, 30% of the population in Malmö had been born abroad.The same figures for Sweden and for the two largest municipalities (Stockholm and Gothenburg) were 15 respective 22% (Statistics Sweden).In the Norwegian study, the immigrant population enclosed 24% of the population (Statistics Norway).Thus, we have a suitable background to further analyze the immigrant population.
The aim of the study was to estimate the prevalence of MS among first-generation immigrants in the City of Malmö and to investigate whether there were differences in disease severity between the immigrants and the native population and to provide a base for further follow-up studies.

Study area and population
Malmö municipality, or the City of Malmö is a Swedish municipality located in the county of Skåne in the south-west of Sweden at 55 • 35` N latitude having a population of 298,963 persons on the prevalence day, 31 December 2010, in an area of 335 km 2 (Statistics Sweden; www.scb.se).The study population included all persons with MS (pwMS) living in Malmö on prevalence day (n = 397) identified through a previous Fig. 1.Flowchart of the distribution of the study population according to country background.Non-Western: Asia (n= 4), Africa (n= 3), Middle East (n=22) and Central and South America (n=1).Middle East: Lebanon, Turkey, Bahrain, Egypt, Kuwait, Syria, Israel, the Palestinian territory, Qatar, Yemen, Iraq, Iran, Jordan, Saudi Arabia, and the United Arab Emirates.survey in the area [25].

Case definition
The term country background was applied for the person's country of birth and categorized into groups (Fig. 1).A native Swedish was defined as a person with country background in Sweden with at least one parent born in Sweden.First-generation immigrants were defined as people who were born abroad by two foreign-born parents, and who later immigrated to Sweden.No information about the country of birth of the parents was available besides that they were not born in Sweden and the same country background as their offspring was assumed.
First-generation immigrants were further defined as Western or non-Western.A Western immigrant was defined as a person with country background in Europe, North America, or Oceania.Individuals from Norway, Denmark and Iceland were excluded from the Western group since they were considered closely related to the Swedish population.Together with the native Swedish group they were defined as the Scandinavian group.A non-Western immigrant was defined as a person with country background in Asia, Africa, the Middle East or South and Central America.
The concept of onset adjusted prevalence and the 2010 McDonald´s diagnostic criteria for MS at the time for case ascertainment were applied [26,27].The pattern and course of the disease was defined according to Lublin [28].Further details about methodology and case ascertainment have been presented previously [25].
To estimate the prevalence in the different immigrant groups, the number of prevalent cases in each group was divided by the number of individuals in the respective country background group who were alive and residing in the City of Malmö on prevalence day.

Case ascertainment 2.3.1. Medical chart review
The medical chart review included assessment of the year of onset, year of diagnosis, onset symptoms, magnetic resonance imaging, oligoclonal bands in the cerebrospinal fluid, the Expanded Disability Status Scale (EDSS) [29] and disease-modifying therapy (DMT).DMT included beta interferon, glatiramer acetate, natalizumab and other treatments (laquinimod, linomide, azathioprine, mitoxantrone).Interferon beta and glatiramer acetate were considered first-line DMT whereas natalizumab was classified as second-line.Onset symptoms were classified using a predefined list of symptoms developed by Poser [30].The onset symptoms were also grouped based on anatomical regions: optic nerve, brainstem/cerebellum, spinal cord, cerebral hemispheres and unclear.
To assess the disability progression rate, we used the Multiple Sclerosis Severity Score (MSSS), which adjusts the EDSS for disease duration from symptom onset.[31].For individuals with disease duration longer than 30 years (n = 61), MSSS was calculated based on 30 years duration.For individuals with year of onset not available (n = 5), MSSS was calculated based on the year of diagnosis.The registered EDSS had been performed between 1980 and 2018.To make the groups more comparable, the calculation of MSSS was restricted to individuals with EDSS performed between 2005 and 2015 (n= 231).The majority of the EDSS (98%) were performed by a neurologist, with 68% performed by an MS-specialist (40% of those by the first author).

The Swedish total population register (TPR)
Since 1947, every individual who has resided in Sweden on a permanent basis has been recorded in the TPR kept by Statistics Sweden.Each individual has been assigned a mandatory personal identity number, which allows to follow an individual over time.The TPR provided information about the country of birth and the time for moving or migration for the pwMS but also population data including country background, sex, and age for the total Malmö population.No data about the length of stay in Sweden for the general population was available.In order to preserve the integrity of the immigrant population, Statistics Sweden did not allow us to calculate the prevalence for individual countries.

Statistical methods
For statistical calculations SPSS 25.0 and Epitools were used.A significance level of 5% was applied.Age-and sex-specific as well as adjusted prevalence rates according to the 2013 European Standard Population were calculated [32].The Wilson`s score method was used for calculation of 95% confidence intervals for the prevalence.Pearson´s chi-square test or 2-samples Z-test were used when comparing frequencies except for small samples (n < 5) when the Fisher´s test was used.Means from normally distributed variables were compared using a two-sample t-test.A Mann-Whitney test was used when comparing skewed continuous variables.For median values, the interquartile range (IQR) was calculated.Differences in MSSS between the groups, adjusted for confounders (sex, age at onset and phenotype) were analysed using multiple linear regression analyses.

Ethics
The study was approved by the local ethics committee of Lund University, Lund (Dnr 2013/890) and the National Board of Health and Welfare (Dnr 10576/2014).

Prevalence
As presented previously, a total of 397 MS individuals were resident in the City of Malmö on prevalence day, 31 December 2010.The total population was 298,963, resulting in a crude prevalence of 133/100,000 [25].The age-adjusted prevalence was 139/100,000 (95% CI, 126-153).For the first-generation immigrants, 80 pwMS were identified in a population of 80,069, yielding a crude prevalence of 100/100,000 (95% CI, 80-124) and an adjusted figure of 79/100,000 (95% CI 62-101) (Tables 1 and 2).The MS prevalence in first-generation immigrants was lower compared to the Scandinavian population of Malmö (p < 0.010 for both crude and adjusted prevalence).
Of the pwMS, 291 of a population of 188,986 had Scandinavian background, resulting in a prevalence of 154/100,000 (95% CI, 137-173).There were 50 pwMS with Western background, of a population of 40,575, yielding a prevalence of 123/100,000 (95% CI, 94-162).For the non-Western group, a total of 30 pwMS were identified in a population of 39,494, yielding a prevalence of 76/100,000 (95% CI, 53-108).The crude and the adjusted prevalence figures are shown in Table 1.The stratified figures are shown in Table 3.
Most of the patients in the non-Western group came from the Middle East (Fig. 1), from Iran (n=10) and Iraq (n=6).In the Western group, 86% came from Central or South Europe.No more information about the country of birth of individual patients is provided with respect to their right to privacy.

Clinical findings
The Scandinavians were older than both immigrant's groups at prevalence day and had a longer disease duration (Table 4, which summarizes the demographic and clinical data).
Overall, the non-Western immigrants were younger at onset than Scandinavians, even when the analysis was restricted to individuals with onset after migration to Sweden (median age 26, IQR 21-31, p < 0.010).When looking into the subgroup with relapsing onset, the non-Western immigrants had also a lower age at onset than the Scandinavians: 26 (IQR 22-30) vs. 30 (IQR 24-37), p = 0.010.The proportion of pwMS with progressive onset was slightly lower in the Scandinavian group (8.9%) than among the immigrants (10% in both groups), which difference was not statistically significant.The median age at disease onset for those with progressive onset was 44 (IQR 37-49) in the Scandinavian group and 33 in both immigrants' groups (IQR 29-49 in Western immigrants and 31-33 in non-Western), which difference was not statistically significant.The sex ratio (female to male) in the primary progressive population was 0.5 among Scandinavians and non-Western immigrants, and 0.7 in Western immigrants.The same figures for those with relapsing onset were 2.4 and 3.2 for the Scandinavians respective Western and non-Western (data not shown).
No differences about diagnosis delay or the time between the diagnosis and the treatment start were found.Of the non-Western patients, 70% had received any medication for MS, and 43% were still receiving first-line treatment at prevalence day.This was significantly higher than for the Scandinavian patients, 49% of whom had received any DMT, whereas 25% were still on first-line therapy (p = 0.030 for both comparisons).The Scandinavian patients had a statistically significant longer time on first DMT than the non-Western immigrants (p= 0.010).
Table 5 describes the distribution of onset symptoms among individuals with relapsing debut.Most of them had a monofocal onset, no statistically significant differences between the groups were found.We further analysed the onset symptoms according to anatomical regions: The most common clinical presentation among non-Western immigrants was brainstem/cerebellar symptoms (48%, p= 0.010 when compared to the Scandinavians).Optic neuritis was the most frequent clinical presentation at onset among both Scandinavians (27%) and Western immigrants (32%).Spinal cord symptoms were the second most frequent clinical presentation in all groups (Data not shown).
The mean MSSS value among Scandinavians was 4.2 (SD 3.5), whereas the figures in the immigrant group were 4.6 (SD 3.3) and 5.2 (SD 3.7) among Westerns respectively non-Westerns, which differences were not statistically significant.When adjusted for gender, age at onset and initial disease course, the mean MSSS difference between non-Western and Scandinavian pwMS was 1,74 (95% CI 0.18-3.30,p = 0.030) (Table 6).

Discussion
We found that both Western and non-Western immigrants had lower MS prevalence than the Scandinavian population.Overall, firstgeneration immigrants had a lower prevalence than the Scandinavian population of Malmö.Our study also showed that non-Western immigrants had a more severe disease course than Scandinavians.Our observations on disability are consistent with previous studies from France, Norway, and USA [19][20][21][22][23].To our knowledge, this is the first time disease disability in immigrants has been studied in Sweden.We found no differences on the diagnosis delay or the time from diagnosis to treatment initiation.
The prevalence rate in the Middle East group was higher than the figures provided in the Atlas of MS from 2013, where most countries in that region had an estimated prevalence of 20-60/100,000.A previous  study of the MS prevalence in 2005 among migrants from the Middle East in Oslo, Norway, based on 14 patients, showed a prevalence of 85/ 100,000 (95% CI, 50-143).These patients came mainly from Iran [33].Interestingly, most of the patients in our Middle East group were also born in Iran.High MS prevalence among Iranian-born Swedish residents has been previously reported [11].Similarly, higher MS-risk was found among first generation women from Iran in a recent Swedish survey [34].Unfortunately, it was not possible to calculate the specific prevalence for individuals born in Iran, which prevented further comparisons.Genome-wide association studies have identified more than 200 genetic risk variants for MS, such as polymorphisms in genes encoding human leukocyte antigen (HLA), with the main risk being associated to HLA-DRB1*15 [35].Most of these studies have been performed on European ancestry populations [36].However, an association between the HLA-DRB1*15 allele and MS risk in the Iranian population has also been described in some studies [37,38].In addition, sun exposure has been significantly related with a decreased risk of MS in Iran [39].Together with previous studies from Norway [33] and Sweden [10,11,34], the present results might suggest that the Iranian-born population bear genetic factors, that in combination with environmental factors in the host country, may contribute to their susceptibility to MS.
As expected, there was a female preponderance in all the groups and among those with relapsing onset.However, we observed a reversed gender bias among the PP population, like the male preponderance founded in a recent European multicenter survey [5].
Older age at onset has been associated with faster disability progression in many studies, even when individuals with PPMS are excluded.Other putative predictors of time to disability are polysymptomatic onset, time to clinically definite disease (CD) and the localization of the first relapse [40].Our study showed an increased disability progression in non-Western immigrants despite an earlier onset, similar time to CD, mainly monofocal onset and no diagnosis or treatment delay compared to Scandinavians.A large study conducted in France with patients identified through the Lyons Multiple Sclerosis Cohort found a favourable effect on prognosis of an optic neuritis onset and an intermediate effect of brainstem onset [41].Similarly, a study from Barcelona showed an association between optic neuritis as disease presentation and slower progression to an EDSS score of 3.0 [42].Of note, optic neuritis was the most common clinical presentation among Scandinavians, whereas the non-Western group had more often a brainstem/cerebellum presentation.
No diagnosis or treatment delay was identified, either when comparing Scandinavians with Western or with non-Western immigrants, which indicate equal access to our health care system.Yet, the use of any DMT among the non-Western immigrants was higher than among the Scandinavians which further confirm the availability of medical services and may denote a more aggressive disease.However, more patients in the non-Western group were still receiving first-line DMT compared to the Scandinavians, while the use of second-line therapy (ever and as first choice) was negligible.In Sweden, natalizumab was approved as a treatment for MS in 2006 but the introduction of the drug in our department was delayed due to economical restrictions.The small number of patients who had received natalizumab on prevalence day may be explained by these financial issues, but it raises the question whether this also reflects a higher threshold for switching therapy in non-Western patients.As previously discussed, a complex interplay of genetic susceptibility and environmental factors is the hallmark of MS etiology and may account for differences in prevalence among different ethnic groups [1].Whether this interplay may also account for differences in disease disability remains unclear.Increased disease disability when moving from a low/medium risk area to a high risk area suggests a key role of environmental factors.Reduced sun exposure when migrating could be one of these factors.Non-Western immigrants showed a more severe disability progression in both our study and in the Oslo survey despite different clinical characteristics (younger age at onset in our study, more often multifocal onset, and faster diagnosis in the Oslo study).Both Norway and Sweden are countries at higher latitudes and a reduction of sun exposure in non-Western migrants can be presumed.However, a recent study from The Middle East North Africa (MENA) region, indicated a more aggressive and disabling MS course in the MENA region compared with the West [43].This was a registry-based study with no unified definitions of clinical phenotypes or disability progression.Besides, DMT are not universal available in the MENA region and the results should thus be interpreted with caution.However, a recently published review concluded that increased disease severity is indeed observed among Black/African and maybe also Hispanic/Latinx individuals with MS, and that this may relate to heightened humoral responses [24].This suggests that the pattern of disability is connate to certain ethnic groups and needs to be studied further.
Strengths of the present study are the availability of linked-register data including country of birth, the rigorous case validation process including medical record review and the high proportion of EDSS performed by an MS specialist.
There are limitations of our study.First, the small numerator, which specially makes interpretations difficult regarding Africa, Asia and Central and South America.One might also speculate whether the immigrant population has a higher threshold for seeking medical care for milder symptoms.If so, it could explain the lower prevalence and the pattern of disability.Second, the lack of data on socioeconomic status.Finally, the selective nature of migration.Immigrants might not be representative of the general population of their country of origin since they are usually younger, healthier and have better socioeconomic status.Nevertheless, and particularly regarding disease disability, there is a consistent pattern from studies in different populations at different times.Taking together these observations with those from the native population in the MENA region it can be argue that the results are more likely to be reliable.
In conclusion, we found a lower prevalence among Western and non-Western first-generation immigrants compared to the Scandinavian population and a more severe disease in non-Western immigrants than in Scandinavians but possibly a higher prevalence compared with their country of origin.Whether genetic susceptibility and/or environmental factors after migration are responsible for the higher disease disability remains unknown.Since the treatment landscape has changed significantly in the last two decades, further studies with current epidemiological data in a larger population are needed.

Funding
This study was in part supported by an unconditional research grant from the Promobilia Foundation (A22065), the Elsa Schmitz Foundation and Neuro Sweden (F2022-0085).

Table 1
Crude and adjusted prevalence of multiple sclerosis in Scandinavians and immigrants on 31 December 2010 in the City of Malmö.

Table 2
Crude and adjusted prevalence of multiple sclerosis among non-Western immigrants on 31 December 2010 in the City of Malmö per 100,000.

Table 3
Prevalence on 31 December 2010 of multiple sclerosis in the City of Malmö per 100,000 population by age, sex, and country background.

Table 4
Clinical characteristics of multiple sclerosis in Scandinavians and immigrants.

Table 6
Multiple Sclerosis Severity Score: results from the multiple linear regression analysis.Covariates: sex, age at onset, phenotype (relapsing/progressive onset).