[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address 52.200.130.163. Please contact the publisher to request reinstatement.
[Skip to Content Landing]
Figure 1.
Seasonal Differences in the UK Government Office Regions (GORs) (1965-1999)
Seasonal Differences in the UK Government Office Regions (GORs) (1965-1999)

A, Deviance in the mean normalized monthly birth rates for the different GORs (data set C3 used). B, Seasonal differences between Scotland and the rest of the United Kingdom (Walter and Elwood test) (data set C3 used). Data sets are reported in the Table.

Figure 2.
Residual Differences in Multiple Sclerosis (MS) Births per Month
Residual Differences in Multiple Sclerosis (MS) Births per Month

A, Observed to expected MS birth ratios in the UK MS cohort (1938-1980) (data sets MS1-5 and C1-C2 used). B, Observed to expected MS birth ratios in the regional UK MS cohort (1965-1980) (data sets MS1-5 and C3 used). Data sets are reported in the Table. Error bars indicate 95% CI.

Table.  
Multiple Sclerosis and Population Control Data Sets
Multiple Sclerosis and Population Control Data Sets
1.
Ebers  GC.  Environmental factors and multiple sclerosis.  Lancet Neurol. 2008;7(3):268-277.PubMedGoogle ScholarCrossref
2.
Handel  AE, Giovannoni  G, Ebers  GC, Ramagopalan  SV.  Environmental factors and their timing in adult-onset multiple sclerosis.  Nat Rev Neurol. 2010;6(3):156-166.PubMedGoogle ScholarCrossref
3.
Rothwell  P.  Measles epidemicity and seasonality of birth in multiple sclerosis.  J Neurol Neurosurg Psychiatry. 1994;57:1286-1287.Google Scholar
4.
Templer  DI, Trent  NH, Spencer  DA,  et al.  Season of birth in multiple sclerosis.  Acta Neurol Scand. 1992;85(2):107-109.PubMedGoogle ScholarCrossref
5.
Salzer  J, Svenningsson  A, Sundström  P.  Season of birth and multiple sclerosis in Sweden.  Acta Neurol Scand. 2010;121(1):20-23.PubMedGoogle ScholarCrossref
6.
Grytten  N, Torkildsen  Ø, Aarseth  JH,  et al.  Month of birth as a latitude-dependent risk factor for multiple sclerosis in Norway.  Mult Scler. 2013;19(8):1028-1034.PubMedGoogle ScholarCrossref
7.
Saastamoinen  K-P, Auvinen  M-K, Tienari  PJ.  Month of birth is associated with multiple sclerosis but not with HLA-DR15 in Finland.  Mult Scler. 2012;18(5):563-568.PubMedGoogle ScholarCrossref
8.
Willer  CJ, Dyment  DA, Sadovnick  AD, Rothwell  PM, Murray  TJ, Ebers  GC; Canadian Collaborative Study Group.  Timing of birth and risk of multiple sclerosis: population based study.  BMJ. 2005;330(7483):120.PubMedGoogle ScholarCrossref
9.
Sotgiu  S, Pugliatti  M, Sotgiu  MA,  et al.  Seasonal fluctuation of multiple sclerosis births in Sardinia.  J Neurol. 2006;253(1):38-44.PubMedGoogle ScholarCrossref
10.
Staples  J, Ponsonby  AL, Lim  L.  Low maternal exposure to ultraviolet radiation in pregnancy, month of birth, and risk of multiple sclerosis in offspring: longitudinal analysis.  BMJ. 2010;340:c1640.PubMedGoogle Scholar
11.
Dobson  R, Giovannoni  G, Ramagopalan  S.  The month of birth effect in multiple sclerosis: systematic review, meta-analysis and effect of latitude.  J Neurol Neurosurg Psychiatry. 2013;84(4):427-432.PubMedGoogle ScholarCrossref
12.
Rothwell  PM, Staines  A, Smail  P, Wadsworth  E, McKinney  P.  Seasonality of birth of patients with childhood diabetes in Britain.  BMJ. 1996;312(7044):1456-1457.PubMedGoogle ScholarCrossref
13.
Samuelsson  U, Johansson  C, Ludvigsson  J.  Month of birth and risk of developing insulin dependent diabetes in south east Sweden.  Arch Dis Child. 1999;81(2):143-146.PubMedGoogle ScholarCrossref
14.
Cooper  JD, Smyth  DJ, Walker  NM,  et al.  Inherited variation in vitamin D genes is associated with predisposition to autoimmune disease type 1 diabetes.  Diabetes. 2011;60(5):1624-1631.PubMedGoogle ScholarCrossref
15.
Ramagopalan  SV, Maugeri  NJ, Handunnetthi  L,  et al.  Expression of the multiple sclerosis–associated MHC class II allele HLA-DRB1*1501 is regulated by vitamin D.  PLoS Genet. 2009;5(2):e1000369.PubMedGoogle ScholarCrossref
16.
Simpson  S  Jr, 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):1132-1141.PubMedGoogle ScholarCrossref
17.
Staples  JA, Ponsonby  A-L, Lim  LL-Y, McMichael  AJ.  Ecologic analysis of some immune-related disorders, including type 1 diabetes, in Australia: latitude, regional ultraviolet radiation, and disease prevalence.  Environ Health Perspect. 2003;111(4):518-523.PubMedGoogle ScholarCrossref
18.
Mohr  SB, Garland  CF, Gorham  ED, Garland  FC.  The association between ultraviolet B irradiance, vitamin D status and incidence rates of type 1 diabetes in 51 regions worldwide.  Diabetologia. 2008;51(8):1391-1398.PubMedGoogle ScholarCrossref
19.
Orton  S-M, Wald  L, Confavreux  C,  et al.  Association of UV radiation with multiple sclerosis prevalence and sex ratio in France.  Neurology. 2011;76(5):425-431.PubMedGoogle ScholarCrossref
20.
Rothwell  PM, Gutnikov  SA, McKinney  PA, Schober  E, Ionescu-Tirgoviste  C, Neu  A; European Diabetes Study Group.  Seasonality of birth in children with diabetes in Europe: multicentre cohort study.  BMJ. 1999;319(7214):887-888.PubMedGoogle ScholarCrossref
21.
Fiddes  B, Wason  J, Kemppinen  A, Ban  M, Compston  A, Sawcer  S.  Confounding underlies the apparent month of birth effect in multiple sclerosis.  Ann Neurol. 2013;73(6):714-720.PubMedGoogle ScholarCrossref
22.
Torkildsen  O, Aarseth  J, Benjaminsen  E,  et al.  Month of birth and risk of multiple sclerosis: confounding and adjustments.  Ann Clin Transl Neurol. 2014;1(2):141-144.PubMedGoogle ScholarCrossref
23.
Palace  J, Duddy  M, Bregenzer  T,  et al.  Effectiveness and cost-effectiveness of interferon beta and glatiramer acetate in the UK Multiple Sclerosis Risk Sharing Scheme at 6 years: a clinical cohort study with natural history comparator.  Lancet Neurol. 2015;14(5):497-505.PubMedGoogle ScholarCrossref
24.
Rothwell  PM, Charlton  D.  High incidence and prevalence of multiple sclerosis in south east Scotland: evidence of a genetic predisposition.  J Neurol Neurosurg Psychiatry. 1998;64(6):730-735.PubMedGoogle ScholarCrossref
25.
Walter  SD, Elwood  JM.  A test for seasonality of events with a variable population at risk.  Br J Prev Soc Med. 1975;29(1):18-21.PubMedGoogle Scholar
26.
Lam  DA, Miron  JA.  Seasonality of births in human populations.  Soc Biol. 1991;38(1-2):51-78.PubMedGoogle Scholar
27.
Bobak  M, Gjonca  A.  The seasonality of live birth is strongly influenced by socio-demographic factors.  Hum Reprod. 2001;16(7):1512-1517.PubMedGoogle ScholarCrossref
28.
Walter  SD.  The power of a test for seasonality.  Br J Prev Soc Med. 1977;31(2):137-140.PubMedGoogle Scholar
29.
Disanto  G, Chaplin  G, Morahan  JM,  et al.  Month of birth, vitamin D and risk of immune-mediated disease: a case control study.  BMC Med. 2012;10:69.PubMedGoogle ScholarCrossref
30.
Holick  MF, Smith  E, Pincus  S.  Skin as the site of vitamin D synthesis and target tissue for 1,25-dihydroxyvitamin D3: use of calcitriol (1,25-dihydroxyvitamin D3) for treatment of psoriasis.  Arch Dermatol. 1987;123(12):1677-1683a.PubMedGoogle ScholarCrossref
31.
Salzer  J, Hallmans  G, Nyström  M, Stenlund  H, Wadell  G, Sundström  P.  Vitamin D as a protective factor in multiple sclerosis.  Neurology. 2012;79(21):2140-2145.PubMedGoogle ScholarCrossref
32.
Ascherio  A, Munger  KL, White  R,  et al.  Vitamin D as an early predictor of multiple sclerosis activity and progression.  JAMA Neurol. 2014;71(3):306-314.PubMedGoogle ScholarCrossref
33.
Munger  KL, Levin  LI, Hollis  BW, Howard  NS, Ascherio  A.  Serum 25-hydroxyvitamin D levels and risk of multiple sclerosis.  JAMA. 2006;296(23):2832-2838.PubMedGoogle ScholarCrossref
34.
Sørensen  IM, Joner  G, Jenum  PA, Eskild  A, Torjesen  PA, Stene  LC.  Maternal serum levels of 25-hydroxy-vitamin D during pregnancy and risk of type 1 diabetes in the offspring.  Diabetes. 2012;61(1):175-178.PubMedGoogle ScholarCrossref
35.
Ford  DV, Jones  KH, Middleton  RM,  et al.  The feasibility of collecting information from people with multiple sclerosis for the UK MS Register via a web portal: characterising a cohort of people with MS.  BMC Med Inform Decis Mak. 2012;12:73.PubMedGoogle ScholarCrossref
36.
Elian  M, Nightingale  S, Dean  G.  Multiple sclerosis among United Kingdom–born children of immigrants from the Indian subcontinent, Africa and the West Indies.  J Neurol Neurosurg Psychiatry. 1990;53(10):906-911.Google ScholarCrossref
37.
Bell  M, Blake  M, Boyle  P,  et al.  Cross-national comparison of internal migration: issues and measures.  J R Stat Soc A. 2002;165(pt 3):435-464.Google ScholarCrossref
38.
Mackenzie  IS, Morant  SV, Bloomfield  GA, MacDonald  TM, O’Riordan  J.  Incidence and prevalence of multiple sclerosis in the UK 1990-2010: a descriptive study in the General Practice Research Database.  J Neurol Neurosurg Psychiatry. 2014;85(1):76-84.PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    1 Comment for this article
    Replicate this study in the Southern Hemisphere
    Charles A. Pilcher MD FACEP | Physician, EvergreenHealth, Kirkland, WA, USA
    The obvious next step is to replicate this study in the Southern Hemisphere.
    CONFLICT OF INTEREST: None Reported
    Original Investigation
    August 2016

    Time- and Region-Specific Season of Birth Effects in Multiple Sclerosis in the United Kingdom

    Author Affiliations
    • 1Nuffield Department of Clinical Neurosciences, University of Oxford, John Radcliffe Hospital, Oxford, England
    • 2The Townsville Hospital, Queensland, Australia
    • 3University of Nottingham, Nottingham, England
    • 4Barts and The London School of Medicine and Dentistry, London, England
    • 5South Eastern Health and Social Care Trust, Belfast, Northern Ireland
    • 6Belfast Trust, Belfast, Northern Ireland
    • 7Imperial College London, London, England
    • 8Cardiff University, Cardiff, Wales
    • 9University of St Andrews, St Andrews, Scotland
     

    Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.

    JAMA Neurol. 2016;73(8):954-960. doi:10.1001/jamaneurol.2016.1463
    Abstract

    Importance  The reports of seasonal variation in the births of people who later develop multiple sclerosis (MS) have been challenged and attributed to the background pattern in the general population, resulting in a false association.

    Objective  To study the seasonality of MS births after adjusting for temporal and regional confounding factors.

    Design, Setting, and Participants  A study was conducted using case-control data from 8 MS-specialized centers from the United Kingdom, MS cases from a population-based study in the Lothian and Border regions of Scotland, and death records from the UK Registrar General. Participants included 21 138 patients with MS and control data from the UK Office of National Statistics and the UK government office regions. The seasonality of MS births was evaluated using the Walter and Elwood test, after adjusting for temporal and regional variations in the live births of the UK population. The study was conducted from January 16, 2014, to September 2, 2015.

    Main Outcomes and Measures  Diagnosis of multiple sclerosis.

    Results  Analysis of the general population indicated that seasonal differences are present across time and region in the United Kingdom, with both factors contributing to the monthly distribution of live births. We were able to demonstrate that, when adjusting for the temporal and regional variations in the live births of the UK population, there was a significant season of birth effect in patients with MS, with an increased risk of disease in the peak month (April) compared with the trough month (November) (odds ratio, 1.24; 95% CI, 1.10-1.41) and 15.68% fewer people who developed MS being born in November (observed to expected birth ratio, 0.840; 95% CI, 0.76-0.92).

    Conclusions and Relevance  Season of birth is a risk factor for MS in the United Kingdom and cannot be attributed to the background pattern in the general population. The reasons for the variations in birth rates in the general population are unclear, but not taking them into consideration could lead to false-positive associations.

    Introduction

    The cause of multiple sclerosis (MS) remains elusive, but environmental factors are thought to be important.1 Environmental exposures might contribute to MS susceptibility acting at different periods.2 Seasonal variation in the births of people who later develop MS has been reported3-11 in different populations with a spring peak and autumn nadir. If this effect is real, early environmental influences may be acting before the disease is clinically evident, and further interrogation of this observation may give clues as to the cause of MS. The same seasonal pattern has also been reported12,13 in childhood-onset type 1 diabetes, and vitamin D and UV radiation have been observed14-19 to modify the risk of disease in both diabetes and MS. However, although the seasonality effect of births in childhood-onset type 1 diabetes is well accepted following appropriately controlled studies,12,20 the validity in MS has been challenged and attributed to the background pattern in the general population.21,22 Adjusting for temporal and geographical variations indicated that the apparent seasonal patterns for month of birth suggested to be specific for MS are expected by chance alone; therefore, previous claims21 for an association of MS with month of birth were probably false-positives. Thus, it is unclear whether a seasonal birth effect exists in MS, and adequately controlled comparisons using the background general population birth rate patterns for the same time and place are required.

    In both MS8 and childhood-onset type 1 diabetes,20 the largest seasonal effects on birth patterns have been observed in the United Kingdom, particularly in Scotland. However, previous studies8,21,22 on MS have been underpowered or lacked appropriate controls to reliably quantify regional and latitudinal effects within the United Kingdom. We therefore studied more than 20 000 UK patients with MS, with the primary objective of determining the month of birth effect with appropriate adjustments for general population time of birth patterns by region.

    Box Section Ref ID

    Key Points

    • Question Is there a month of birth effect in multiple sclerosis (MS) after the appropriate adjustments for general population time of birth patterns by region in the United Kingdom?

    • Findings In this case-control study that included more than 20 000 patients with MS in the United Kingdom and time- and region-matched controls, there was a significant season of birth effect in patients with MS.

    • Meaning This study provides data consistent with the hypothesis that very early environmental influences contribute to the risk of developing MS.

    Methods
    MS Data

    Month and year of birth of 31 806 UK individuals with a diagnosis of MS were derived from 5 main sources (Table). Most of the data were supplied by 8 specialist regional MS centers (Oxford University Hospitals National Health Service Trust, University of Nottingham, Cardiff University, Plymouth University, Imperial College London, The Walton Centre Liverpool, Barts and The London School of Medicine and Dentistry, and Northern Ireland Neurology Service). Each regional center had a local research ethics agreement for database storage and analysis of deidentified patient data. The other sources included data on individuals collected as part of the UK Risk Sharing Scheme provided by the Multiple Sclerosis Trust, which was established by the Department of Health in 2002 to monitor a cohort of patients with MS to ensure cost-effective provision of interferon beta and glatiramer acetate. Eligibility included all UK patients with MS who met the Association of British Neurologist criteria for use of these drugs.23 In addition, population-based MS incidence and prevalence studies in the Lothian and Border regions of Scotland24 and death records from the UK Registrar General were included. Individuals with MS who were born before 1938 (9304 [29.3%]) were excluded because annual live births of the general population were not recorded monthly in England and Wales before that year; therefore, adequate controls were not available. Those with MS who were born after 1980 (1364 [4.29%]) were excluded because persons born in that period (especially males, who tend to develop MS at a later stage) will not have lived through their entire disease risk period; thus, post-1980 control data were not used to adjust the MS data. Each regional MS specialist center (Oxford University Hospitals NHS Trust, University of Nottingham, Cardiff University, Plymouth University, Imperial College London, The Walton Centre Liverpool, Barts and The London School of Medicine and Dentistry, and Northern Ireland Neurology Service) had local ethics agreement for the storage and analysis of deidentified patient data through their local research and development departments. Ethical approval for the Risk Sharing Scheme was given by the South East Medical Research Ethics Committee. The general population data are available to the public. The study was conducted from January 16, 2014, to September 2, 2015.

    General Population Control Data

    The control data were obtained from the UK Office of National Statistics and were composed of 3 data sets (Table). Two population data sets covered England and Wales (1941-2000) and Scotland (1941-2000), including monthly annual live births for each year. The third source included regional data from the UK government office regions (1965-1999), which was made up of monthly annual live births for each region. Regional birth data were not collected before 1965 in the United Kingdom. The latter data were obtained to compare with an MS cohort from the same period matched on birth location.

    Normalization of Control Data

    We adjusted for the seasonality of live births in the general population by calculating the normalized birth rate per month for each year (r). This adjustment was done by constructing pseudocohorts of births adjusted for the different number of days per month, including leap years, assuming no seasonal variation (eMethods in the Supplement). The deviance (excess or deficit) in the normalized birth rate per month was calculated as D = 100 × (r − 1) and represents the percentage difference of the monthly annual live births compared with the pseudocohorts previously described (eMethods in the Supplement).

    Examination for Potential Temporal Effect

    We examined temporal changes in the seasonality of births in the general population by using the UK (England, Wales, and Scotland) monthly annual live births data sets from the period 1938-2000. Deviances in the monthly rates over time were plotted together with their respective trend lines.

    Examination for Potential Regional Effect

    First, we compared the monthly distribution of live births for England and Wales in 1938-2000 with the distribution in Scotland using the Walter and Elwood test.25 This test was specifically designed to investigate the seasonality of events with a variable population at risk. The strength of this test is that it does not assume any specific expected seasonal pattern, such as the sinusoidal pattern, that is the basis of other tests of seasonality. Rather, the Walter and Elwood test evaluates whether the center of gravity (centroid) of the “clock face” of monthly births differs significantly between 2 groups (eFigure 1 in the Supplement). The test only assumes that any seasonal trend will be asymmetrical: any excess of observed births will tend to cluster in adjacent months and with a roughly annual periodicity, resulting in a difference between case and control populations in their centers of gravity on the clock face, the significance of which can then be calculated using a single χ2 test. Therefore, this test avoids the issue of multiple comparisons by not using month-specific χ2 tests of association for individual months.

    Second, we performed a regional analysis by using the monthly annual live-births figures available from different UK government office regions in 1965-1999. We looked for regional differences in the seasonality of births by comparing the monthly annual live births of Scotland (ie, the region with the most extreme latitude) with all other UK regions (Walter and Elwood test) and by plotting the mean normalized monthly birth rates of the regions against latitude.

    Comparison of the MS and Control Cohorts
    Assessing Seasonality After Adjusting for Temporal Effect

    We calculated the MS-expected births for each month after adjusting the live births in the control population to the relative frequency of MS births per year (1938-1980) and for country of origin. The expected numbers were compared with those observed using the Walter and Elwood test. The peak-trough month ratio was calculated as an approximate measure of the amplitude of seasonality and is given as an odds ratio (95% CI).

    Assessing Seasonality After Adjusting for Temporal and Regional Effect

    We calculated the MS-expected births for each month after adjusting the live births from the UK government office regions to the relative frequency of MS births per region (and per year to allow for a temporal effect). The Walter and Elwood test was used to compare the expected numbers with the observed numbers. The peak-trough month ratio was calculated as an approximate measure of the amplitude of seasonality. Because the regional control data set covered live births during 1965-1999 and the regional MS data (obtained from the specialist MS centers and the Multiple Sclerosis Trust) covered births during 1899-1980, we selected data from 1965-1999.

    Statistical Analysis

    Data were originally registered in a database (Microsoft Excel 2010; Microsoft Corp). Statistical analysis was performed using SPSS, version 19 (IBM). The Walter and Elwood test was performed according to the authors’ instructions.25P < .05 was used as evidence for a statistically significant difference.

    Results
    Temporal and Regional Effects In Control Data

    Examination of a regional effect on control birth rates showed seasonal differences in the births of the general population between England and Wales compared with Scotland during most individual years (eFigure 2A in the Supplement) and decades (eFigure 2B in the Supplement) in 1938-2000. The regional analysis using the mean data available within 1965-1999 demonstrated significant differences between Scotland and several southern regions of England and Wales; for example, there was a 5.31% excess in births in the South West region in May compared with Scotland (Figure 1). These differences were greater in regions with milder latitudes (eFigure 3 in the Supplement).

    Examination of the temporal effect in the UK general population showed a correlation between the deviance in the normalized birth rates and the year of birth except for January, June, and December, when the tendency reached a plateau (eFigure 4 in the Supplement). Therefore, seasonal differences in case-control studies might change depending on the time-frame population used as controls (eFigure 5 in the Supplement).

    Control vs MS Monthly Birth Rates
    UK MS Cohort (1938-1980)

    A seasonal effect was found in the UK MS group (n = 21 138) compared with the UK general population from the same period after adjusting for the relative frequency of MS births per year and for country of origin (P < .001), with a peak-trough amplitude of 1.17 (95% CI, 1.09-1.25) compared with controls. The peak-trough months were April (observed to expected birth ratio, 1.07; 95% CI, 1.02-1.11), with 6.77% more MS births, and November (observed to expected birth ratio, 0.91; 95% CI, 0.87-0.95), with 9.01% fewer MS births than expected (Figure 2A). When grouping the data into quartiles, the seasonal effect persisted (eFigure 6 in the Supplement).

    Regional MS Cohort (1965-1980)

    There remained a seasonal effect in the regional MS group (n = 6372) compared with the regional controls after adjusting for the relative frequency of MS births per year and region (P < .001), with a peak-trough amplitude of 1.24 (95% CI, 1.10-1.41) compared with controls. This effect was particularly marked in November (observed to expected birth ratio, 0.840; 95% CI, 0.76-0.92), with 15.68% fewer MS births than expected (Figure 2B). When the data were grouped into quartiles, the seasonal effect persisted (eFigure 7 in the Supplement).

    Discussion

    The findings of this study appear to confirm that, after making the appropriate corrections for regional origin and year of birth, the month of birth effect in development of MS in the United Kingdom remains significant. We have shown that seasonal differences in population birth rates are present across time and regions in the United Kingdom, with both factors contributing to the monthly distribution of births in the general population. Therefore, these confounders should be considered when studying the seasonality of diseases.

    In most populations, birth rates vary by season, although not in an identical manner.26 The causes of these seasonal variations are not fully understood, although cultural, environmental, and socioeconomic factors could have an important effect.27 It is also not clear what drives the temporal and regional trends in the seasonality of births in the general population. It is interesting that the seasonality in the control population appears to have declined over time (eFigure 4 in the Supplement), although control data from the past decade suggest that the seasonal effect has reversed (eFigure 8 in the Supplement). These findings highlight the need to use reliable controls in association studies that are matched on year of birth and regional origin. The size of the regions chosen should also be considered since the use of smaller regions can be useful in detecting regional MS risk microvariations.19 However, the use of small regions is often limited by the availability of data.

    The validity of the season of birth effect in MS has been questioned21 on the basis that some previous studies used unreliable controls unmatched for time and region of birth. The other problem with previous studies has been a lack of statistical power to reliably quantify modest seasonality of birth.3,8 To avoid both of these pitfalls, we used the largest UK MS cohort to date and adjusted the control data sets according to the relative frequency of MS births per time and region of birth. The power of the Walter and Elwood test to detect a seasonal variation of 5% amplitude or more for a given size of approximately 25 000 MS cases would be as high as 99% (P < .05).28 However, the power of this test will, in general, be higher than that of the usual χ2 test for heterogeneity of rates between the number of time periods.

    We found a consistent seasonal pattern in a large MS cohort weighted by country of birth and year of birth. This pattern was also confirmed in a cohort of patients with MS weighted by regional origin (regional effect) and year of birth (temporal effect). These results support the findings of previous smaller studies in Australia (1524 patients with MS)10 and the Italian region of Sardinia (810 patients with MS)9 that used adequate controls.

    The seasonal pattern of MS is shared by other immune-mediated diseases, especially type 1 (insulin-dependent) diabetes.29 In the United Kingdom, type 1 diabetes has a similarity of season of birth with MS in relation to the same peak-trough months, latitude trends in seasonality, and overall epidemiology.20 This similarity supports the idea of a common factor responsible for the environmental effects seen in both conditions. One of the main candidates is sunlight exposure since UV radiation is the main determinant of vitamin D concentration,30 and numerous studies have pointed out that poor vitamin D status is associated with an increased risk of MS.1,31-33 In addition, the vitamin D hypothesis in type 1 diabetes is closer to being established and widely accepted.34 However, there are other possible environmental factors that could act in the perinatal period, including viral infections during pregnancy and dietary differences related to season.2 Environmental exposures could also have an effect during childhood and adulthood.

    The limitations of our study include the lack of information on socioeconomic class, race, culture, and other factors, which could be biases; the differences in data acquisition; and the controls not matched for sex, although there seem to be no appreciable sex differences in population birth rates.21 Other limitations are the accuracy of death certificates for MS because 100% specificity cannot be guaranteed and the possibility of a low number of duplicate individuals in the data sets owing to the deidentified data. We were not able to exclude patients with MS who were not born in the United Kingdom because this information was not systematically collected. However, data from the UK MS register35 and migration studies36 suggest that such persons represent a very small proportion of the sample. In addition, we have not allowed for individuals with MS who move from their birth regions, although internal migration studies in the United Kingdom suggest a small effect.37 The fact that patients with MS are identified only after diagnosis could also represent a bias for those born in the 1970s (especially males, who tend to develop the disease later and have not lived through their entire disease risk period),38 increasing the proportion of females with relapsing-remitting MS and an earlier age of onset. This limitation does not apply to individuals born earlier, and therefore will not have an effect throughout the entire MS cohort. However, these factors are unlikely to produce a systematic bias that would cause a seasonal effect once time and region are controlled for. Some of the limitations may have introduced noise into the analysis and thus could have diluted our findings.

    Conclusions

    The findings of our study demonstrate that the month of birth effect is a risk factor for MS in the UK population, with a pattern similar to that reported for type 1 diabetes. This study provides data consistent with the hypothesis that very early environmental influences contribute to the risk of developing MS. Regional origin and year of birth influence the seasonality of population birth rates and therefore should be considered when studying the seasonality of births in disease groups. Further studies are required to determine the cause of the month of birth effect in MS in the United Kingdom. The symmetrical deviations could imply a natural cycle; therefore, studying factors such as the variation in sunlight hours across regions of the United Kingdom would be of great interest.

    Back to top
    Article Information

    Corresponding Author: Jacqueline Palace, FRCP, DM, Nuffield Department of Clinical Neurosciences, John Radcliffe Hospital, Headley Way, Level 3, West Wing, Headington OX3 9DU, England (jacqueline.palace@ndcn.ox.ac.uk).

    Accepted for Publication: April 5, 2016.

    Published Online: June 27, 2016. doi:10.1001/jamaneurol.2016.1463.

    Author Contributions: Dr Palace had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

    Study concept and design: Rodríguez Cruz, Matthews, Boggild, Constantinescu, Giovannoni, Oppenheimer, Rothwell, Palace.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Rodríguez Cruz, Rothwell, Palace.

    Critical revision of the manuscript for important intellectual content: Matthews, Boggild, Cavey, Constantinescu, Evangelou, Giovannoni, Gray, Hawkins, Nicholas, Oppenheimer, Robertson, Zajicek, Rothwell.

    Statistical analysis: Rodríguez Cruz, Matthews, Oppenheimer, Rothwell.

    Administrative, technical, or material support: Boggild, Cavey, Evangelou, Hawkins, Nicholas, Robertson, Zajicek, Palace.

    Study supervision: Boggild, Hawkins, Zajicek, Palace.

    Conflict of Interest Disclosures: Dr Constantinescu has received research support, travel support for meetings, and consultancy fees from Biogen Idec, Bayer-Schering, Merck Serono, Novartis, Sanofi-Pasteur MSD, Morphosys, and Teva. Dr Giovannoni has received compensation for serving as a consultant or speaker for or has received research support from AbbVie, Bayer-Schering Healthcare, Biogen Idec, Canbex, Eisai, Elan, Five Prime Therapeutics, Genzyme, Genentech, GlaxoSmithKline, Ironwood Pharmaceuticals, Merck-Serono, Novartis, Pfizer, Roche, Sanofi, Synthon BV, Teva Pharmaceutical Industries, UCB, and Vertex Pharmaceuticals. Ms Gray has received unrestricted educational grants from Biogen Idec, Merck Serono, and Novartis; honoraria as a consultant on scientific advisory boards for Genzyme, Biogen Idec, Merck Serono, and Novartis; and has participated in clinical trials by Biogen Idec and Merck Serono. Her institution has received research support from Merck Serono. Dr Palace is partially funded by highly specialized services to run a national congenital myasthenia service and a neuromyelitis optica service. She has received support for scientific meetings and honorariums for advisory work from Merck Serono, Biogen Idec, Novartis, Teva Pharmaceutical Industries, Chugai Pharma, and Bayer-Schering and unrestricted grants from Merck Serono, Novartis, Biogen Idec, and Bayer-Schering. Her hospital trust receives funds for her role as clinical lead for the UK Risk Sharing Scheme, and she has received grants from the Multiple Sclerosis Society and Guthie Jackson Foundation for unrelated research studies. She is a board member for the charitable European MS foundation, “The Charcot Foundation,” and is on the steering committee for a European collaborative MS imaging group, Magnetic Resonance Imaging in MS.

    Additional Contributions: We are grateful to the UK multiple sclerosis specialist centers, UK Multiple Sclerosis Trust, and the UK Office for National Statistics for providing the required data to conduct this study.

    References
    1.
    Ebers  GC.  Environmental factors and multiple sclerosis.  Lancet Neurol. 2008;7(3):268-277.PubMedGoogle ScholarCrossref
    2.
    Handel  AE, Giovannoni  G, Ebers  GC, Ramagopalan  SV.  Environmental factors and their timing in adult-onset multiple sclerosis.  Nat Rev Neurol. 2010;6(3):156-166.PubMedGoogle ScholarCrossref
    3.
    Rothwell  P.  Measles epidemicity and seasonality of birth in multiple sclerosis.  J Neurol Neurosurg Psychiatry. 1994;57:1286-1287.Google Scholar
    4.
    Templer  DI, Trent  NH, Spencer  DA,  et al.  Season of birth in multiple sclerosis.  Acta Neurol Scand. 1992;85(2):107-109.PubMedGoogle ScholarCrossref
    5.
    Salzer  J, Svenningsson  A, Sundström  P.  Season of birth and multiple sclerosis in Sweden.  Acta Neurol Scand. 2010;121(1):20-23.PubMedGoogle ScholarCrossref
    6.
    Grytten  N, Torkildsen  Ø, Aarseth  JH,  et al.  Month of birth as a latitude-dependent risk factor for multiple sclerosis in Norway.  Mult Scler. 2013;19(8):1028-1034.PubMedGoogle ScholarCrossref
    7.
    Saastamoinen  K-P, Auvinen  M-K, Tienari  PJ.  Month of birth is associated with multiple sclerosis but not with HLA-DR15 in Finland.  Mult Scler. 2012;18(5):563-568.PubMedGoogle ScholarCrossref
    8.
    Willer  CJ, Dyment  DA, Sadovnick  AD, Rothwell  PM, Murray  TJ, Ebers  GC; Canadian Collaborative Study Group.  Timing of birth and risk of multiple sclerosis: population based study.  BMJ. 2005;330(7483):120.PubMedGoogle ScholarCrossref
    9.
    Sotgiu  S, Pugliatti  M, Sotgiu  MA,  et al.  Seasonal fluctuation of multiple sclerosis births in Sardinia.  J Neurol. 2006;253(1):38-44.PubMedGoogle ScholarCrossref
    10.
    Staples  J, Ponsonby  AL, Lim  L.  Low maternal exposure to ultraviolet radiation in pregnancy, month of birth, and risk of multiple sclerosis in offspring: longitudinal analysis.  BMJ. 2010;340:c1640.PubMedGoogle Scholar
    11.
    Dobson  R, Giovannoni  G, Ramagopalan  S.  The month of birth effect in multiple sclerosis: systematic review, meta-analysis and effect of latitude.  J Neurol Neurosurg Psychiatry. 2013;84(4):427-432.PubMedGoogle ScholarCrossref
    12.
    Rothwell  PM, Staines  A, Smail  P, Wadsworth  E, McKinney  P.  Seasonality of birth of patients with childhood diabetes in Britain.  BMJ. 1996;312(7044):1456-1457.PubMedGoogle ScholarCrossref
    13.
    Samuelsson  U, Johansson  C, Ludvigsson  J.  Month of birth and risk of developing insulin dependent diabetes in south east Sweden.  Arch Dis Child. 1999;81(2):143-146.PubMedGoogle ScholarCrossref
    14.
    Cooper  JD, Smyth  DJ, Walker  NM,  et al.  Inherited variation in vitamin D genes is associated with predisposition to autoimmune disease type 1 diabetes.  Diabetes. 2011;60(5):1624-1631.PubMedGoogle ScholarCrossref
    15.
    Ramagopalan  SV, Maugeri  NJ, Handunnetthi  L,  et al.  Expression of the multiple sclerosis–associated MHC class II allele HLA-DRB1*1501 is regulated by vitamin D.  PLoS Genet. 2009;5(2):e1000369.PubMedGoogle ScholarCrossref
    16.
    Simpson  S  Jr, 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):1132-1141.PubMedGoogle ScholarCrossref
    17.
    Staples  JA, Ponsonby  A-L, Lim  LL-Y, McMichael  AJ.  Ecologic analysis of some immune-related disorders, including type 1 diabetes, in Australia: latitude, regional ultraviolet radiation, and disease prevalence.  Environ Health Perspect. 2003;111(4):518-523.PubMedGoogle ScholarCrossref
    18.
    Mohr  SB, Garland  CF, Gorham  ED, Garland  FC.  The association between ultraviolet B irradiance, vitamin D status and incidence rates of type 1 diabetes in 51 regions worldwide.  Diabetologia. 2008;51(8):1391-1398.PubMedGoogle ScholarCrossref
    19.
    Orton  S-M, Wald  L, Confavreux  C,  et al.  Association of UV radiation with multiple sclerosis prevalence and sex ratio in France.  Neurology. 2011;76(5):425-431.PubMedGoogle ScholarCrossref
    20.
    Rothwell  PM, Gutnikov  SA, McKinney  PA, Schober  E, Ionescu-Tirgoviste  C, Neu  A; European Diabetes Study Group.  Seasonality of birth in children with diabetes in Europe: multicentre cohort study.  BMJ. 1999;319(7214):887-888.PubMedGoogle ScholarCrossref
    21.
    Fiddes  B, Wason  J, Kemppinen  A, Ban  M, Compston  A, Sawcer  S.  Confounding underlies the apparent month of birth effect in multiple sclerosis.  Ann Neurol. 2013;73(6):714-720.PubMedGoogle ScholarCrossref
    22.
    Torkildsen  O, Aarseth  J, Benjaminsen  E,  et al.  Month of birth and risk of multiple sclerosis: confounding and adjustments.  Ann Clin Transl Neurol. 2014;1(2):141-144.PubMedGoogle ScholarCrossref
    23.
    Palace  J, Duddy  M, Bregenzer  T,  et al.  Effectiveness and cost-effectiveness of interferon beta and glatiramer acetate in the UK Multiple Sclerosis Risk Sharing Scheme at 6 years: a clinical cohort study with natural history comparator.  Lancet Neurol. 2015;14(5):497-505.PubMedGoogle ScholarCrossref
    24.
    Rothwell  PM, Charlton  D.  High incidence and prevalence of multiple sclerosis in south east Scotland: evidence of a genetic predisposition.  J Neurol Neurosurg Psychiatry. 1998;64(6):730-735.PubMedGoogle ScholarCrossref
    25.
    Walter  SD, Elwood  JM.  A test for seasonality of events with a variable population at risk.  Br J Prev Soc Med. 1975;29(1):18-21.PubMedGoogle Scholar
    26.
    Lam  DA, Miron  JA.  Seasonality of births in human populations.  Soc Biol. 1991;38(1-2):51-78.PubMedGoogle Scholar
    27.
    Bobak  M, Gjonca  A.  The seasonality of live birth is strongly influenced by socio-demographic factors.  Hum Reprod. 2001;16(7):1512-1517.PubMedGoogle ScholarCrossref
    28.
    Walter  SD.  The power of a test for seasonality.  Br J Prev Soc Med. 1977;31(2):137-140.PubMedGoogle Scholar
    29.
    Disanto  G, Chaplin  G, Morahan  JM,  et al.  Month of birth, vitamin D and risk of immune-mediated disease: a case control study.  BMC Med. 2012;10:69.PubMedGoogle ScholarCrossref
    30.
    Holick  MF, Smith  E, Pincus  S.  Skin as the site of vitamin D synthesis and target tissue for 1,25-dihydroxyvitamin D3: use of calcitriol (1,25-dihydroxyvitamin D3) for treatment of psoriasis.  Arch Dermatol. 1987;123(12):1677-1683a.PubMedGoogle ScholarCrossref
    31.
    Salzer  J, Hallmans  G, Nyström  M, Stenlund  H, Wadell  G, Sundström  P.  Vitamin D as a protective factor in multiple sclerosis.  Neurology. 2012;79(21):2140-2145.PubMedGoogle ScholarCrossref
    32.
    Ascherio  A, Munger  KL, White  R,  et al.  Vitamin D as an early predictor of multiple sclerosis activity and progression.  JAMA Neurol. 2014;71(3):306-314.PubMedGoogle ScholarCrossref
    33.
    Munger  KL, Levin  LI, Hollis  BW, Howard  NS, Ascherio  A.  Serum 25-hydroxyvitamin D levels and risk of multiple sclerosis.  JAMA. 2006;296(23):2832-2838.PubMedGoogle ScholarCrossref
    34.
    Sørensen  IM, Joner  G, Jenum  PA, Eskild  A, Torjesen  PA, Stene  LC.  Maternal serum levels of 25-hydroxy-vitamin D during pregnancy and risk of type 1 diabetes in the offspring.  Diabetes. 2012;61(1):175-178.PubMedGoogle ScholarCrossref
    35.
    Ford  DV, Jones  KH, Middleton  RM,  et al.  The feasibility of collecting information from people with multiple sclerosis for the UK MS Register via a web portal: characterising a cohort of people with MS.  BMC Med Inform Decis Mak. 2012;12:73.PubMedGoogle ScholarCrossref
    36.
    Elian  M, Nightingale  S, Dean  G.  Multiple sclerosis among United Kingdom–born children of immigrants from the Indian subcontinent, Africa and the West Indies.  J Neurol Neurosurg Psychiatry. 1990;53(10):906-911.Google ScholarCrossref
    37.
    Bell  M, Blake  M, Boyle  P,  et al.  Cross-national comparison of internal migration: issues and measures.  J R Stat Soc A. 2002;165(pt 3):435-464.Google ScholarCrossref
    38.
    Mackenzie  IS, Morant  SV, Bloomfield  GA, MacDonald  TM, O’Riordan  J.  Incidence and prevalence of multiple sclerosis in the UK 1990-2010: a descriptive study in the General Practice Research Database.  J Neurol Neurosurg Psychiatry. 2014;85(1):76-84.PubMedGoogle ScholarCrossref
    ×