Risk factors for multiple sclerosis in the Northern Isles of Scotland
Item statusRestricted Access
Embargo end date30/06/2019
Weiss, Emily Margaret
This thesis looks at risk factors for multiple sclerosis (MS), a chronic, degenerative autoimmune disease which is usually diagnosed between the ages of 20 and 50 years. It is estimated to affect over 100,000 people in the UK. The research setting was Orkney and Shetland, two archipelagos situated north of mainland Scotland, and both of which have very high MS prevalence as do other countries at high latitudes. I examine genetic and environmental risk factors in Orkney and Shetland using multiple methods over four studies. I also review the vitamin D and UV exposure literatures as these are risk factors pertinent to MS in Orkney and Shetland. After devoting three chapters to introducing the purpose of the thesis, MS, and Orkney and Shetland, in the fourth chapter, I aim to establish whether the birthplace of cases show any spatial, temporal, or spatiotemporal clustering. Evidence of these kinds of clustering may indicate that there are environmental risk factors present in some areas or that were present over particular periods, which raise risk of developing MS. Although I find statistically significant temporal, spatial, and spatiotemporal clustering in Orkney, and a spatial cluster in Shetland, for multiple reasons these results need to be interpreted with caution. I conclude that the clusters are very likely to be artefacts. Furthermore, there are multiple possible alternative explanations for such clusters that could not be explored by the available data. Chapter 5 examines the heritability of MS in Orkney and Shetland to estimate the proportion of phenotypic variance attributable to additive genetic effects. I also look at the birthplaces of ancestors of cases and controls to see if any locations contribute a greater amount of ancestral DNA to the gene pool of modern MS cases, which I term ‘genetic clustering’. In Orkney I obtained a heritability estimate of 0.36 (95% CI -0.26, 0.98); in Shetland this estimate was 0.20 (95% CI -1.88, 2.28). These modest estimates are consistent with the heritability literature. The genetic clustering analyses highlight two Orkney registration districts, Kirkwall and Westray, which earlier studies identified as areas of MS clustering. I also identify three Shetland registration districts, however these locations had not shown any evidence of clustering in earlier studies. Again, I advise caution in interpreting results, particularly as all the error bars across registration districts overlap. Chapter 6 presents a scoping review to map the literature and identify evidence of an association between vitamin D and UV exposure with MS. In methodically searching the literature, I identify a large and heterogeneous evidence base comprising multiple observational, intervention, and genetic studies. Overall, many studies support an association between vitamin D deficiency and MS. There is also evidence for an association between UV exposure and MS, although UV exposure is considerably less explored than vitamin D. I finally identify gaps in the literature and make suggestions for future research. In Chapter 7 I aim to compare vitamin D levels in Orkney and mainland Scotland, and establish the determinants of vitamin D status in Orkney. I firstly compare mean vitamin D and prevalence of deficiency in cross-sectional data from studies in Orkney and mainland Scotland. I secondly use multivariable regression to identify factors associated with vitamin D levels in Orkney. I find that mean (standard deviation) vitamin D is significantly higher in Orkney compared to mainland Scotland (35.3 (18.0) and 31.7 (21.2), respectively), and prevalence of severe deficiency is lower in Orkney (6.6% to 16.2% p = 1.1 x 10-15). Factors associated with higher vitamin D in Orkney include older age, farming occupations and foreign holidays. I conclude that although mean vitamin D levels are higher in Orkney compared to mainland Scotland, there is substantial variation within the Orkney population which may influence MS risk. Chapter 8 examines the correlates and determinants of UVB exposure in Shetland. I firstly construct correlation matrices to visualise how 1) personal characteristics such as sex, occupation, and skin type, 2) physical activity, and 3) body weight and fat, correlate with UVB exposure. I then use multivariable regression to identify factors associated with UVB exposure in Shetland. I run two multivariable models. The first includes the full sample size where activity data were measured by questionnaires. The second includes both questionnaire physical activity data and step-count data from pedometers, however as only a subset of participants had been supplied with pedometers, this analysis comprises a smaller sample size. I find that the amount of skin exposed was most strongly correlated with UVB exposure. Step count and activity minutes were also moderately positively correlated, and indoor occupations moderately negatively correlated, with UVB exposure. The regression analysis using the full sample with questionnaire activity data found that factors associated with greater UVB exposure were age and ambient UVB, while working indoors was significantly associated with lower UVB exposure. The model including the pedometer data found that found that age, total steps, and the amount of ambient UVB were significantly associated with greater UVB exposure. I conclude that atmospheric conditions, working outdoors and older age are important factors in UVB exposure in Shetland. It remains to be seen how UVB exposure translates to vitamin D levels in Shetland. I found evidence for environmental and genetic risk factors for MS in Orkney and Shetland. The two environmental risk factors, vitamin D deficiency and reduced UV exposure, are more likely to affect the younger population who are still within their lifetime risk of developing MS.