Non-parametric analysis of seasonality in birth and multiple sclerosis risk in second generation of migrants in Kuwait

Multiple sclerosis (MS) is a highly debilitating immune mediated disorder of the central nervous system and is often progressive and fatal. MS etiology seems to be determined by environmental factors and genetic predisposition [1]. The contribution of each factor in MS etiology may vary in different geographic areas resulting into atypical spatial patterns of its incidence across the globe [2]. Hitherto known spatial pattern of MS reflects the highest prevalence, morbidity and mortality rates in the temperate zones of both northern and southern hemispheres and with decreasing rates in the subtropics and tropics. For instance, Europe, Canada and the northern parts of United States are high risk areas, whereas the Asia, Africa and West Indies are regarded as low risk areas [3]. So far, no specific exogenous or genetic basis for the spatial pattern has been identified, but it is speculated that some climatologic condition influences the frequency of the disease. It is unknown whether this effect is a direct on the patient or an indirect effect on the animal or plant life in his environment [1],[4].

Studies on migratory populations have indicated that the MS risk among those migrating from a high to a low risk area exceeds that of the population into which they have immigrated. Furthermore, those who migrate from a high-risk area to a low-risk area keep their high risk unless they migrate below the age of sixteen [5]-[7]. There are inconsistent reports regarding MS risk among immigrants from low risk areas such as Asia, East Africa and Caribbean to high MS risk areas. Additionally, as noted earlier, MS risk depends on age at the time of migration [8]. Reportedly, MS was relatively uncommon among West Indian and Asian immigrants to UK [8]-[10]. However, children of Asian and West Indian immigrants born in England and Wales had MS prevalence similar to that in the general population of England and Wales [11],[12].

Based on Kurtzke classification, Arabian Gulf region is located in a low-risk zone for MS [13]. However, recent studies suggest a moderate-to-high MS prevalence (range: 31–55 cases × 10⁻⁵ individuals) with an increase in incidence in recent years [14],[15]. Arabian Gulf countries including Kuwait have a large population of migrant workers from Asian and African countries. Prior to first Gulf War in 1990, MS prevalence in Kuwait was significantly higher among non-Kuwaiti (mainly Palestinians) (23.8 × 10⁻⁵) than Kuwaiti (9.5 ×10⁻⁵) [16]. However, the composition of population in Kuwait radically changed due to sociopolitical reasons since the first Gulf War in 1990. There was an influx of migrant workers from South and South-east Asia and other Arab countries replacing a pre-war a major expatriate group of Palestinians [15]. Furthermore, during the post- Gulf War era, the MS prevalence in Kuwait increased from 6.7 × 10⁻⁵ in 1993 to 14.8 ×10⁻⁵ in 2000. Moreover, in a complete reversal of the pattern observed before 1990, this difference was more pronounced among Kuwaiti (31.2 × 10⁻⁵) than non-Kuwaiti (5.6 × 10⁻⁵) migrant workers. Therefore, in a geographic area with previously low MS prevalence, the local environment may be responsible for the dramatic change in MS risk [15]. Seasonal pattern in MS births has been consistent of several studies conducted in a number of regions and countries [17]. Month of birth is a surrogate for continuous environmental trait which might be present during the entire year and peak at certain period rather than being present only on month or season [1]. Previously in Kuwait, we demonstrated birth of month effect on the MS risk among the Kuwaiti born in Kuwait using parametric statistical models. However, similar relationship did not hold for non-Kuwaiti born in Kuwait [18]. This study assessed the overall MS incidence and evaluated the month of birth effect on MS risk among non-nationals born, lived and diagnosed with MS in Kuwait from January 1950 to April 2013 using non-parametric analysis of the data from Kuwait National MS Registry.

During the study period, non-nationals born in Kuwait had an overall risk of MS births (per 10⁵ non-Kuwaiti births in general population) as 23.8 (95% CI: 20.8 – 27.0). Gender-specific risk of MS births (per 10⁵ non-Kuwaiti births by gender in general population) showed non-Kuwaiti female had statistically significant (p = 0.003) increased risk (28.6; 95% CI: 24.2 – 33.7) than non-Kuwaiti males (18.7; 95% CI: 15.1–23.0) (Table 1). Non-Kuwaiti females born and lived in Kuwait were 1.5 times more likely to have been diagnosed with MS than non-Kuwait males born and lived in Kuwait (Relative risk =1.5; 95% CI: 1.2 – 2.0). The month-specific distribution of non-Kuwaiti MS births compared with non-Kuwaiti births in general population did not differ significantly (χ ² goodness-of-fit test statistic = 9.51, p = 0.575). Of rank-sums for successive 6-month segments based on the Hewitt’s non-parametric test, from September through-February period showed an evidence of slight but statistically non-significant (p = 0.090) increased tendency of non-Kuwaiti MS births (Tables 2 and 3).

In this evaluation, the proportion of migrants MS births over the study period was statistically significantly higher than previously reported Kuwaiti MS births [18]. Non-parametric analysis revealed slight but statistically non-significant increased tendency of migrants’ MS births from September through February. There are inconsistent reports on the question of MS risk in migrant populations. The MS risk among migrants seems to be dependent not only on a sufficiency of people who change their residence from one risk area to another but also on ages at immigration and their length of stay (exposure period) in the new land. Additionally, this relationship of geography and MS risk is further confounded by predilection for MS regardless of region [25],[26]. There is evidence from mortality data that moving from low-risk to high-risk MS regions does increase the risk of dying from MS [27]. Unpublished data also showed an increased MS risk among southern-born US-veterans who moved to North of US and Vietnamese who moved to France, though these results were considered non-confirmatory [27]. UK born-children of West Indian immigrants had the same magnitude of probable MS incidence and prevalence as was among Irish population in Northern Ireland. This finding indicated that although genetic factors play role but there is strong evidence that the cause of disease is mainly environmental and therefore preventable [28].

In line with earlier studies investigating the genetic and spatial epidemiology of this complex disease, more recent studies have highlighted how MS arises from a combination of genetic susceptibility and environmental exposures acting from gestation to early adulthood. For example, in a recent meta-analysis it has been argued that latitude-dependent month of birth effect on MS risk is likely due to ultraviolet light exposure and maternal vitamin D level [29]. Conversely, Fiddes et al. reasoned-out that confounding underlies this apparent association of month of birth and MS risk. The authors claimed that using national birth statistics as controls may generate false-positive association of MS with month of birth [30]. However, a limitation of aforementioned study is that the researchers did not analyze the original studies that highlighted the month of birth effect in MS [30]. Instead, they performed a separate study on national un-weighted births statistics, thereby showing potential sources of bias in such studies [30],[31]. Nevertheless, in search of causal relationships, season of birth, maternal Vitamin D deficiency, Epstein Barr virus infection, and smoking behaviour are strongly implicated and seem to influence genetic predisposition to MS. Additionally, these factors appear to act synergistically and the risk of MS in individuals exposed to more than one factor combines multiplicatively [1],[4],[31],[32]. Further research is needed to fully understand the environmental risk factors for MS and in this regard longitudinal studies of at-risk populations should be considered [32]. Knowledge of such risk factors may help design preventive strategies to minimize MS risk.

The proportion of migrants’ MS births (per 10⁵ non-national births in general population) over the study period was 23.8 (95% CI: 20.8 – 27.0), which was statistically significantly higher than previously reported Kuwaiti MS births (16.2; 95% CI: 15.1 – 17.4) in Kuwait. Non-parametric analysis revealed slight but statistically non-significant increased tendency of migrants’ MS births from September through February. Knowledge of MS risk factors and how and when they act among genetically vulnerable individuals from gestation to early adulthood will help design prevention strategies.