You are here
Perinatal characteristics and obstetric complications in mothers with multiple sclerosis: Record-linkage study
Multiple Sclerosis and Related Disorders, Volume 12, February 2017, Pages 4–8
Multiple sclerosis (MS) predominantly onsets in women of reproductive age. The possibility of adverse obstetric and perinatal outcomes is a likely source of concern to pregnant women with MS and their clinicians. We aimed to compare the characteristics of the pregnancies of mothers with or without MS.
The historical Oxford Record Linkage Study specialised maternity dataset (850,000 people, 1970–1989), with record-linkage between mother and baby, was analysed. The dataset was linked to any prior recorded day-case or inpatient hospital admission episodes back to 1963. The file of mothers’ records was searched for a record of MS in either the maternity admission or in a previous admission. The pregnancies and babies of mothers with MS were compared with those of mothers without MS.
There were 181 pregnancies and babies born to 98 mothers with MS. These were compared with 244,573 pregnancies and babies of 124,830 mothers without MS. There was a significant social class gradient with a higher than expected number of cases of MS in the least deprived social classes. Mothers with MS tended to be lighter than other mothers. There were no significant associations between MS and mothers’ marital status, history of smoking during the pregnancy, parity, pre-eclampsia, ABO blood group or rhesus group. There were no significant associations with babies’ birth weight, and no significant associations with gestational age, being small for gestational age, Caesarian delivery, or forceps delivery. There were no stillbirths, no neonatal deaths, and no postneonatal deaths in the babies born to mothers with MS.
We hope that our findings will add to the available literature in addressing the understandable anxieties of young women with MS, and reassure them that the characteristics of their pregnancies are generally normal.
- Women with MS do not appear to have an increased risk of neonatal complications.
- Women with MS do not appear to have an increased risk of obstetric complications.
- Babies born to mothers with MS were similar to other babies.
Keywords: Multiple sclerosis, Maternal, Perinatal, Risk factors, Epidemiology.
Multiple sclerosis (MS) is a complex neurological disease which predominantly onsets in women (sex ratio 3:1) aged 20–40 years, i.e. of reproductive age (Koch-Henriksen and Sorensen, 2010). The possibility of adverse obstetric and perinatal outcomes is a likely concern both to clinicians caring for pregnant women with MS and to the women themselves. Information available to reassure patients has been limited by relatively small studies where findings are inconclusive, and from studies where the reporting of obstetric complications has been a secondary outcome to the study of any influence of pregnancy on the clinical course of MS (Orvieto et al, 1999 and Worthingon et al, 1994). Where larger studies have now been done, findings have been inconsistent. For example, reports have been conflicting in whether offspring of mothers with MS are more likely to be pre-term, small for gestational age, and whether operative or instrumental interventions during delivery are more frequently used than in other mothers (Dahl et al, 2005, Mueller et al, 2002, Jalkanen et al, 2010, Chen delete 'et al, 2009, and Kelly et al, 2009). We analysed a historical dataset of routinely collected maternity data, enhanced by record linkage between the mother and baby records, with the objective of comparing the characteristics of the pregnancies of mothers with or without MS. Some of the characteristics will reflect those of women with MS irrespective of pregnancy (e.g. social class); others relate to the pregnancy itself (e.g. baby's birth weight and gestational age).
2.1. Data and population
The Oxford Record Linkage Study (ORLS), founded in 1963, was a project run jointly by the National Health Service (NHS) and the University of Oxford in the (former) Oxford NHS Region. It comprised routinely collected data on all hospital admissions and deaths in the area covered by it. The data collection systems were generally similar to those collected routinely on hospital care throughout the NHS in England as Hospital Activity Analysis (HAA) in the 1960s–1980s (HAA is now superceded by the Hospital Episode Statistics system, HES). To avoid duplication, the core HAA and ORLS systems in Oxford were the same data. However, of relevance to this paper, the ORLS was enhanced by additional data fields in its maternity dataset from 1970 to 1989. The enhanced specialist maternity data collection system included information on maternal smoking, individual-level occupational social class of the head of the mother's household, mother's marital status, mother's weight (but not height), breastfeeding, presence or absence of pre-eclampsia, Apgar score and several other data items that were not in standard HAA and are not in its successor in England, maternity HES. In this enhanced data collection system, the maternity data were collected by dedicated maternity clerks trained for the purpose. The clerks interviewed the patients and contributed to, and had full access to, the clinical record working under the supervision of consultant obstetricians. The record of each admission for the mother was linked to all her other admission records in the ORLS from 1963 to 1999. Thus, for example, a mother who gave birth in 1980 had linked records (if she had previous admissions) back to 1963 and forward to 1999.
The population covered by the specialist maternity system was that of Oxfordshire and West Berkshire (population size approximately 850,000 people). The populations covered by data collection for general hospital admissions and for deaths in the ORLS – to which the mothers' and infants' records were linked - were these two districts throughout the ORLS from 1963, plus four other adjacent districts from 1975, and all eight districts of the former Oxford NHS Region from 1991 (population 2.5 million).
In the ORLS maternity dataset, linkage of the mother-to-baby record, for each mother-baby pair, was initiated, as a routine, by the clerks in the maternity hospitals at the time of the birth (see above). A unique ORLS identifier was allocated to the mother, and a separate one to the baby. The mother's identifier was also coded on the baby record; and the baby's identifier was coded on the mother's record. There are some missing data: for example, data on social class and mothers' weight were not collected in every year. The table will show totals for each variable and the reader can see the extent of missing data (which was generally small).
2.2. Analysis: mothers with MS
The file of mothers' records was searched for a record of MS in either the maternity admission or in a previous admission. MS was identified as the International Classification of Diseases (ICD) code 345 in the 7th edition of the ICD, 340 in the 8th and 9th, and G35 in the 10th. The pregnancies and babies of mothers with MS were compared with those of mothers without MS. As indicated below, in most analyses the unit of analysis was the individual pregnancy and baby (counting each baby born to each mother); in a few analyses, notably social class, mother's weight, and blood group, we counted only the first baby of each mother in order to avoid multiple-counting of the mothers' characteristic. Where there is an ‘ordered’ characteristic with more than two values, such as social class (five values, ordered from high to low socio-economic status), we calculated chi-squared statistics for both heterogeneity (shown first in the table) and for trend (shown second).
3.1. Mothers with MS
There were 181 pregnancies and babies born to 98 mothers with MS. These were compared with 244,573 pregnancies and babies of 154,830 mothers without MS. Comparisons are shown in Table 1. At the time of the 181 MS deliveries, 26 babies were born to mothers aged under 25, 134 to mothers aged 25–34, and 21 to mothers aged 35 and over (Table). There was a significant social class gradient with a higher than expected number of cases of MS in social classes 1 and 2 (the least deprived) and a lower than expected number of cases in social classes 4 and 5 (the most deprived; X2(1) for trend of decreasing MS with increasing deprivation =12.6, p<0.01). Mothers with MS tended to be lighter than other mothers: 71% of mothers with MS weighed nine stone or less compared with 57% of mothers without MS (p=0.04, Table 1). There were no significant associations between MS and mothers’ marital status, history of smoking during the pregnancy, parity, pre-eclampsia, ABO blood group or rhesus group. There were no significant associations with babies’ birth weight (a slightly higher percentage of babies of mothers with MS than other babies weighed less than 2000 g but the percentages of these in the MS and non-MS group were both very low at, respectively, 2.8% and 1.4%), and no significant associations with gestational age, being small for gestational age, Caesarian delivery, or forceps delivery. There was a borderline significant difference in presentation at birth between the babies of MS mothers and other babies (respectively, 92.3% and 95.5% were vertex, p=0.06). There was also a borderline significant difference in Apgar scores at one minute: 6.4% of babies of MS mothers were scored 10, compared with 12.8% of other babies. However, the numerically important differences were between the categories of Apgar 9 and Apgar 10: both are comfortably within the normal range and the finding is unlikely to have any clinical importance. There were no differences between the scores of babies of mothers with or without MS at 5 min. The baby was male in 51.4% of births to MS mothers and 51.4% of other births. Mothers with MS were fractionally more likely than other mothers to breastfeed (Table 1), but the difference was not statistically significant. There were no stillbirths, no neonatal deaths, and no postneonatal deaths in the babies born to mothers with MS. In babies of mothers without MS, the stillbirth, neonatal and postneonatal mortality rates were, respectively, 5.22 per 1000 births (based on 1278 stillbirths), 4.14 (based on 1015 deaths) and 3.33 (based on 806 deaths). Applying these rates to the number of babies born of mothers with MS, the ‘expected’ number of stillbirths, neonatal and postneonatal deaths in the MS group would have been, respectively, 0.94, 0.75 and 0.60.
Babies of mothers with multiple sclerosis: perinatal and maternal characteristics.
|Characteristics||Numbers||Percentages||X2 (d.f.)a||P-value||Adjustment for YOB and mothers' age|
|With MS||Without MS||With MS||Without MS||Odds||C.I.|
|Mother's age (years)|
|1 (high status)||18||13837||20.2||13.1||13.4 (4)||0.01||1||1|
|5 (low status)||4472||0.0||4.2||0–0|
|9st (57.2 kg) & under||54||43545||71.1||56.9||11.7 (6)||0.04||1||1|
|10st (63.5 kg)||7||15730||9.2||20.6||1.2 (1)||0.28||0.4||0.2–0.8|
|11st (69.9 kg)||6||8958||7.9||11.7||0.5||0.2–1.2|
|12st (76.2 kg)||6||4058||7.9||5.3||1.2||0.5–2.7|
|13st (82.6 kg)||1||2108||1.3||2.8||0.5||0.1–2.8|
|14st (88.9 kg)||1008||0.0||1.3||0–0|
|15st (95.3 kg)||2||1068||2.6||1.4||1.5||0.4–6.2|
|Cigarettes per day|
|Gestational age (weeks)|
|Birthweight of baby (g)|
|Small for gestational age (<10% decile)|
|ABO blood groupb|
a The first chi-squared statistic shown is that for heterogeneity; the second, where relevant and given, is that for trend.
b Counting each mother once only (based on first birth for each mother's system number).
Our findings are reassuring to women with MS that their pregnancies generally seem to follow a normal course. Specifically, the babies born to mothers with MS were similar to other babies in respect of birth weight, gestational age, lack of intrapartum growth retardation (as measured by light birth weight for gestational age), mode of delivery (notably, Caesarian section rates were no higher than in the pregnancies of mothers without MS); and there were no recorded stillbirths or deaths in infancy in the babies of mothers with MS. These findings are in line with some previous work and, importantly with the results from a large study utilising a British Columbian cohort, which incorporated clinical factors including disease duration of MS, age of MS onset and disease-associated disability, in which maternal MS did not appear to increase risk of adverse perinatal outcomes (van der Kop et al., 2011). Further, in a systematic review and meta-analysis of women with MS and their pregnancies it was concluded that women with MS do not appear to have a significantly increased risk of obstetric or neonatal complications (Finkelsztejn et al., 2011). Of interest, our finding of a higher occurrence of MS in higher social classes contributes to a very inconsistent evidence base with different studies reporting a positive, nil or negative association (Goulden et al., 2015).
The strengths of this study include that it contributes to an increasing literature on pregnancy outcomes in women with MS; and that the data on births were collected at the time of the birth. The fact that the data are linked – mother record to infant record - is a fundamental strength. The latter means that biases like recall bias, which may occur in interview-based case control studies of people with MS and controls, are impossible. A further strength is that the maternity data include items that, even now, are not in modern routine systems like English Hospital Episode Statistics, such as individual-person social class, marital status, maternal smoking and Apgar scores. The obvious limitations are relatively small numbers and the fact that the birth records are not recent. On small numbers, we suggest that the data on the characteristics of mothers with MS, and their pregnancies and infant outcomes, are sufficient to conclude that generally the pregnancies of such mothers are not compromised in substantial ways. We also hope their publication will add to the availability of data for future meta-analyses. On the fact that the data are old, we hope by this publication to stimulate interest in the use of modern electronic health records which could be used to monitor the outcomes of births routinely in women with MS. Another limitation includes the absence of clinical data, other than diagnostic data, and thereby an inability to stratify women by severity of disease or other disease-specific clinical variables.
In conclusion, the principal finding was that the characteristics of the pregnancies were normal. This contributes to a growing evidence base for clinicians to draw upon when addressing the likely anxiety of young women considering pregnancy in the face of chronic neurological disease.
Acknowledgements and funding
The building of the linked datasets, and the development of the analytical software used to study disease associations, was funded by the English National Institute for Health Research (RNC/035/002). This study had no specific funding. The funder had no role in study design, data collection, data analysis, data interpretation, writing of the report or for the decision to submit for publication. The views expressed in the paper do not necessarily reflect those of the funding body.
All authors declare no conflicts of interest. AG, JP, and MJG report no disclosures.
AG and MG proposed the study. AG undertook all the analyses working as a visiting student in the Oxford University Department of Public Health. JP and AG wrote the first draft of the manuscript. MJG is the guarantor of and designed the study. All authors contributed to the interpretation of the data and revision of the manuscript for important intellectual content.
- Chen delete 'et al., 2009et al., 2009 Y.H. Chen, H.L. Lin, H.C. Lin. Does multiple sclerosis increase risk of adverse pregnancy outcomes? A population-based study. Mult. Scler.. 2009;15(5):606-612 Crossref
- Dahl et al., 2005 J. Dahl, K.M. Myhr, A.K. Daltveit, J.M. Hoff, N.E. Gilhus. Pregnancy, delivery, and birth outcome in women with multiple sclerosis. Neurology. 2005;65(12):1961-1963 Crossref
- Finkelsztejn et al., 2011 A. Finkelsztejn, J.B. Brooks, F.M. Paschoal Jr, Y.D. Fragoso. What can we really tell women with multiple sclerosis regarding pregnancy? A systematic review and meta-analysis of the literature. BJOG. 2011;118(7):790-797 Crossref
- Goulden et al., 2015 R. Goulden, T. Ibrahim, C. Wolfson. Is high socioeconomic status a risk factor for multiple sclerosis? A systematic review. Eur. J. Neurol.. 2015;22(6):899-911 Crossref
- Jalkanen et al., 2010 A. Jalkanen, A. Alanen, L. Airas. Finnish multiple sclerosis and pregnancy study group. Pregnancy outcome in women with multiple sclerosis: results from a prospective nationwide study in Finland. Mult. Scler.. 2010;16(8):950-955 Crossref
- Kelly et al., 2009 V.M. Kelly, L.M. Nelson, E.F. Chakravarty. Obstetric outcomes in women with multiple sclerosis and epilepsy. Neurology. 2009;73(22):1831-1836 Crossref
- Koch-Henriksen and Sorensen, 2010 N. Koch-Henriksen, P.S. Sorensen. The changing demographic pattern of multiple sclerosis epidemiology. Lancet Neurol.. 2010;9:520-532 Crossref
- van der Kop et al., 2011 M.L. van der Kop, M.S. Pearce, L. Dahlgren, A. Synnes, D. Sadovnick, A.L. Sayao, H. Tremlett. Neonatal and delivery outcomes in women with multiple sclerosis. Ann. Neurol.. 2011;70(1):41-50 Crossref
- Mueller et al., 2002 B.A. Mueller, J. Zhang, C.W. Critchlow. Birth outcomes and need for hospitalization after delivery amon women with multiple sclerosis. Am. J. Obstet. Gynecol.. 2002;186(3):446-452 Crossref
- Orvieto et al., 1999 R. Orvieto, R. Achiron, Z. Rotstein, S. Noy, I. Bar-Hava, A. Achiron. Pregnancy and multiple sclerosis: a 2-year experience. Eur. J. Obstet. Gynecol. Reprod. Biol.. 1999;82(2):191-194 Crossref
- Worthingon et al., 1994 J. Worthingon, R. Jones, M. Crawford, A. Forti. Pregnancy and multiple sclerosis – a 3-year prospective study. J. Neurol.. 1994;241(4):228-233
Unit of Health-Care Epidemiology, Nuffield Department of Population Health, University of Oxford, Oxford, United Kingdom
⁎ Correspondence to: Prof Michael Goldacre, Unit of Health-Care Epidemiology, Nuffield Department of Population Health, University of Oxford, Old Road Campus, Oxford OX3 7LF, United Kingdom.
© 2016 Elsevier B.V., All rights reserved.