Article Text

Download PDFPDF

Original research
Childbirth delivery mode and the risk of multiple sclerosis: a prospective population-based study
Free
  1. Akash Kapali1,2,
  2. Anne Kjersti Daltveit1,3,
  3. Kjell-Morten Myhr2,4,
  4. Kjetil Bjornevik5,6,
  5. Elisa Baldin7,
  6. Maura Pugliatti8,9,
  7. Trond Riise1,2,
  8. Marianna Cortese6
  1. 1 Department of Global Public Health and Primary Care, University of Bergen, Bergen, Norway
  2. 2 Neuro-SysMed, Department of Neurology, Haukeland University Hospital, Bergen, Norway
  3. 3 Department of Health Registry Research and Development, Norwegian Institute of Public Health, Bergen, Norway
  4. 4 Department of Clinical Medicine, University of Bergen, Bergen, Norway
  5. 5 Department of Epidemiology, Harvard T H Chan School of Public Health, Boston, Massachusetts, USA
  6. 6 Department of Nutrition, Harvard T H Chan School of Public Health, Boston, Massachusetts, USA
  7. 7 Epidemiology and Biostatistics Unit, IRCCS Istituto Delle Scienze Neurologiche di Bologna, Bologna, Emilia-Romagna, Italy
  8. 8 Department of Neuroscience and Rehabilitation, University of Ferrara, Ferrara, Italy
  9. 9 Department of Medicine, McGill University, Montreal, Québec, Canada
  1. Correspondence to Akash Kapali, Department of Global Public Health and Primary Care, University of Bergen, Bergen 5009, Norway; akash.kapali{at}uib.no

Abstract

Background Caesarean section (CS) may affect the risk of developing multiple sclerosis (MS) in the offspring, possibly through changes in gut microbiota composition, but findings from previous studies are inconsistent. We investigated whether birth by CS was associated with the risk of adult-onset MS.

Methods We conducted a prospective population-based cohort study, including all individuals born in Norway between 1967 and 2003, using the Medical Birth Registry of Norway linked with the Norwegian Multiple Sclerosis Registry and Biobank. The follow-up was until 2021. We used multivariable Cox models to estimate HRs for MS risk with 95% CIs.

Results Among 2 046 637 individuals in the cohort, 4954 MS cases were identified. Being born by CS was associated with a modest increase in MS risk (HR=1.18, 95% CI 1.05 to 1.32). In the sibling-matched analysis, we found no association between CS and MS risk. We found an interaction between CS and gestational age (p=0.03): CS was associated with an increased risk of MS in individuals born preterm (HR=1.62, 95% CI 1.18 to 2.24), whereas there was no association in individuals born at term (HR=1.13, 95% CI 0.99 to 1.27). In a subgroup analysis of individuals born in 1988 and onwards, emergency CS was related to an elevated MS risk (HR=1.40, 95% CI 1.07 to 1.83), whereas planned CS was not (HR: 1.10, 95% CI 0.77 to 1.58).

Conclusions CS was associated with a modestly higher risk of developing MS. However, the stronger associations seen in subgroups who likely experienced a more complicated pregnancy/delivery may point to confounding underlying these associations.

  • MULTIPLE SCLEROSIS

Data availability statement

Data may be obtained from a third party and are not publicly available. Data may be obtained from a third party and are not publicly available. Data from the Norwegian Multiple Sclerosis Registry and MBRN are accessible to researchers by application.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • Being born by caesarean section (CS) may delay or disturb the development of the immune system and increase the risk for immune-mediated diseases such as asthma and inflammatory bowel disease. Whether the delivery mode modifies the susceptibility for developing multiple sclerosis (MS) remains unknown, as findings from previous studies are inconsistent.

WHAT THIS STUDY ADDS

  • Birth by CS is associated with a slightly increased risk of MS compared with birth by vaginal delivery. However, the association was driven by individuals born preterm and individuals born by emergency CS who may have experienced a more complicated gestation and delivery and display underlying characteristics that confound the association between CS and MS.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • This prospective cohort study adds to the evidence that delivery by CS is not related to the development of MS.

Introduction

Multiple sclerosis (MS) is a chronic disabling disease of the central nervous system, affecting approximately 2.8 million people worldwide.1 The disease is likely caused by a combination of genetic predisposition and exposure to certain environmental factors. Although infection with the Epstein-Barr virus (EBV) has been found to be the leading cause of MS,2 other environmental factors, such as smoking,3 low vitamin D4 and a high body mass index in early life,5 have consistently been associated with a higher MS risk, and therefore likely play a role considering that most adults are infected with EBV. Thus, the search for other risk factors remains highly relevant.

Being born by caesarean section (CS) has been linked to the risk of MS, but the evidence is inconsistent. In a case–control study from Iran, CS was associated with a significantly higher risk of MS compared with vaginal delivery.6 However, this finding was not replicated in a cohort study from Denmark.7 Still, CS has been associated with an increased risk of other immune-mediated diseases such as asthma8 9 and allergies,10–12 and autoimmune diseases such as inflammatory bowel disease,13 suggesting that the mode of childbirth delivery affects immune mechanisms that may be relevant for MS. Further, as newborns delivered by CS are not exposed to maternal vaginal and faecal bacteria at birth, there is a delayed colonisation pattern of the fetal intestine that can lead long-term alterations in the gut microbiota,14 which has been associated with the risk for MS.15 16

There has been an upward trend in the global CS rate in the past few decades, including in European countries17 and in Norway,18 and at the same time, the incidence of MS has increased.1 The reasons behind the observed increase in the rates of MS are not entirely understood, and it remains unclear whether the mode of childbirth delivery could be a contributing factor.

Therefore, we investigated whether being born by CS was associated with the risk of developing MS in adulthood, using data from a nationwide registry-based cohort of all individuals born between 1967 and 2003 in Norway.

Methods

Study design and population

We conducted a cohort study among all individuals born between 1967 and 2003 in Norway (n=2 133 795), registered in the Medical Birth Registry of Norway (MBRN). By using the personal identification number unique to every Norwegian resident, data from the MBRN were linked to the Norwegian Multiple Sclerosis Registry and Biobank (MS Registry) to identify individuals who later developed MS. Individuals registered in the MBRN from 1967 to 2003 were followed up until 31 December 2021. We excluded individuals who died or emigrated from Norway or had an onset of MS before age 18. We also excluded the MS cases who had neither information on the year of their MS debut nor MS diagnosis. The study population comprised 2 046 637 individuals alive at the age of 18 and living in Norway (figure 1).

Figure 1

Flow diagram of the study population selection. MBRN, Medical Birth Registry of Norway; MS, multiple sclerosis.

Data sources

The MBRN, established in 1967, is a continuously updated mandatory registry that comprises information about all pregnancies in Norway from 12 weeks of gestation, including terminations after 12 weeks. It includes information about the child and parents, information about maternal health before and during the pregnancy, as well as any complications and interventions during the pregnancy or delivery. The attending midwife is responsible for the completion of the notification form at the end of each birth. The registry is routinely linked with the National Population Registry run by Statistics Norway (SN)19 20 to ensure complete coverage.

SN is the national statistical institute of Norway and produces official statistics in the country. SN was used to obtain sociodemographic information such as the educational status of the study participants and their parents, household and personal income, as well as emigration status.21

Information on date of MS symptoms, onset and diagnosis, and initial disease course (relapsing–remitting or progressive MS) was obtained from the MS Registry. Diagnoses are verified by neurologists according to the McDonald criteria. The consent-based coverage of the MS Registry was more than 80% in Norway at the time of linkage in 2021.22 23 While the coverage of the registry continues to increase, there are still some regional differences in the registration, as some neurological departments actively register cases, while some are still establishing registration routines. The coverage is estimated by using the Norwegian National Patient Registry (NPR). NPR is a mandatory registry of given hospital diagnoses according to the International Statistical Classification of Diseases and Related Health Problems, 10th Revision, codes. All MS diagnoses of patients in Norwegian specialist care have been registered in NPR since March 2007. The data regarding MS diagnoses in NPR have previously been validated, with a sensitivity of 97% and a positive predictive value of 92%.24

Exposure and covariates

The exposure of interest was the mode of delivery (vaginal delivery vs CS). Information on the types of CS (planned or emergency) was available for pregnancies from the year 1988. We selected a priori potential confounders for the association between CS and the risk of MS to be tested: sex (male or female), year of birth (5-year intervals), birth weight (<2500, 2500–3999 and ≥4000 g), gestational age (preterm: 20–36 weeks, term: ≥37 weeks), birth order (first born, second or higher born), maternal age at delivery (<25, 25–34 and ≥35 years), parental education when the individual was 16 years old (tertiary education completed, tertiary education started, high school and mandatory education) and household income in quintiles when the individual was 16 years old (only available from 1988 and onwards).

Statistical analyses

We examined the association between the mode of delivery and the risk of MS using Cox proportional hazard models with calendar year as the timescale and reported the results as HRs and 95% CIs. Each cohort member was followed up from the age of 18 years until the end of follow-up on 31 December 2021, the year of the person’s death, the year of emigration or the year of onset of the first symptoms of MS, whichever came first. In multivariable-adjusted models, the HRs were adjusted for sex, year of birth, birth weight, gestational age, birth order, maternal age at delivery and parental education when the individual was 16 years old.

We used robustness estimates of the variance to account for dependencies between maternal siblings. We also performed a sibling-matched Cox regression analysis to account for potential shared confounders among siblings from the same mother. We examined possible interaction between CS and the covariates by introducing an interaction term in the multivariable-adjusted Cox regression model. Furthermore, we ran multivariable-adjusted Cox regression models for types of CS, restricting the year of birth from 1988 to 2003. These multivariable-adjusted models were further adjusted for household income when the individual was 16 years old.

All the tests were two-tailed, and a p value of less than 0.05 was considered statistically significant. All analyses were conducted using STATA V.16, and figures were made in R using the ggplot2 package V.3.4.0.

Results

The study population included 2 046 637 individuals, of whom 9.0% (range: 2% in 1967 to 16% in 2003) were born by CS. The median follow-up time was 16 years (range 0–36 years). During the follow-up, 1486 men and 3468 women were diagnosed with MS. The demographic and clinical characteristics of the study participants are shown in table 1.

Table 1

Baseline characteristics of study participants and their mothers according to mode of delivery

Overall, we observed a statistically significant association between CS and the risk of developing MS (HR=1.18, 95% CI 1.05 to 1.32) in an analysis adjusted for sex, year of birth, birth weight, gestational age, birth order, maternal age at delivery and parental education (figure 2). The HR did not change materially when further adjusting for maternal morbidity (diabetes, preeclampsia, hypertension, rheumatoid arthritis, chronic kidney disease and asthma) (data not shown). In the sibling-matched Cox regression analysis performed to adjust for potential unmeasured confounders shared by siblings, there was no longer an association between CS and the risk of MS compared with the association seen in the overall analysis (HR=1.07, 95% CI 0.81 to 1.41). When we stratified the analysis by sex, we found that CS was associated with a 23% increased risk of MS in men (HR=1.23, 95% CI 1.01 to 1.50) and a 15% increased risk in women (HR=1.15, 95% CI 1.00 to 1.32) (p for interaction, p=0.57) (table 2). There was a statistically significant interaction between CS and gestational age (p=0.03). Among individuals born preterm, delivery by CS was associated with an elevated MS risk later in life compared with vaginal delivery (HR=1.62, 95% CI:1.18 to 2.24), whereas in individuals born at term, there was no association between CS and MS risk (HR=1.13, 95% CI 0.99 to 1.27) (table 3). There was no evidence of an interaction between CS and the other covariates: year of birth, birth weight, birth order, maternal age at delivery or parental education (data not shown).

Table 2

Association between CS and risk of MS by sex., multivariable-adjusted HRs with 95% CIs

Table 3

Association between Cs and risk of MS by gestational age, multivariable-adjusted HRs with 95% CIs

Figure 2

Association between CS and risk of MS, stratified by sex and gestational age. crude and multivariable-adjusted HRs with 95% CIs. CS, caesarean section; MS, multiple sclerosis; PYAR, person-years at risk. Model 1: Crude Cox regression analysis. Model 2: Cox regression analysis adjusted for sex and year of birth. Model 3: Multivariable-adjusted Cox regression analysis adjusted for variables: sex, year of birth, birth weight, gestational age, birth order, maternal age at birth, education of parents. Overall: Risk of MS among individuals born by CS compared with those born by vaginal delivery. Male: Risk of MS among males born by CS compared with males born by vaginal delivery. Female: Risk of MS among females born by CS compared with females born by vaginal delivery. Preterm: Risk of MS among individuals born preterm by CS compared to individuals born preterm by vaginal delivery. Term: Risk of MS among individuals born at term by CS compared to individuals born at term by vaginal delivery.

In a subanalysis, which we restricted to individuals born between 1988 and 2003 and included 826 individuals with MS and 900 755 individuals without MS, we did not find any meaningful difference in the HR when further adjusting for household income when the individual was 16 years old (data not shown). In this same subgroup, emergency CS was related to an elevated MS risk (HR=1.40, 95% CI 1.07 to 1.83), while planned CS was not (HR 1.10, 95% CI 0.77 to 1.58) (figure 3).

Figure 3

Associations between caesarean delivery and risk of MS. Multivariable-adjusted HRs with 95% CIs, restricting year of birth from 1988 to 2003. CS, caesarian section; MS, multiple sclerosis; PYAR, person-years at risk. Model 1: Crude Cox regression analysis. Model 2: Cox regression analysis adjusted for sex and year of birth. Model 3: Multivariable-adjusted Cox regression analysis adjusted for variables: sex, year of birth, birth weight, gestational age, birth order, maternal age at birth, education of parents.

Discussion

In this cohort study of Norwegians followed up for several decades, we found an association suggesting an 18% increased risk of MS among individuals born by CS compared with those born vaginally when several potential confounding factors were considered. However, the association did not persist in the sibling-matched analysis. Furthermore, the association is only present in individuals born preterm and in individuals born by emergency CS. Hence, considering that we found an association in subgroups of individuals who more likely experienced complicated pregnancies and deliveries and who may have inherently different underlying characteristics and risk profiles, the modest association seen between CS and MS may be attributed to unmeasured confounding.

Findings from some previous studies about CS and the risk of MS have been inconsistent. Our findings contrast with those reported in a study from Iran, in which the authors reported a 2.5-fold increased risk of MS (95% CI 1.43 to 4.41) among individuals born by CS compared with those born by vaginal delivery. This retrospective case–control study included patients with MS (n=449) and their healthy siblings (n=900) and relied on retrospective self-reports of the mode of delivery, gestational age and birth order after MS was diagnosed, which are more prone to bias.6 Our findings are in line with the only other prospective study conducted on this topic, a nationwide cohort study from Denmark, which included 945 individuals with MS born from 1977 to 2007. The authors reported no significant association between CS and the risk of MS (risk ratio (RR)=1.17, 95% CI 0.92 to 1.46) with similar estimates as in the present study, when adjusting for birth order, gestational age, birth weight, age and calendar period.7

Whether CS increases the risk for immune-mediated disorders such as MS has been of particular interest as the proportion born by CS has increased markedly in many countries worldwide over the past decades.17 18 Evidence from previous studies suggests that CS is associated with a range of diseases, such as asthma,8 9 inflammatory bowel disease,13 allergies10–12 and diabetes.25 The hypotheses arose because of potential differences in the microbiota in babies born by CS. During vaginal delivery, the child is exposed to the vaginal and intestinal microbiota while passing through the birth canal, while a child born by CS is mainly exposed to skin bacteria.26 27 Findings from recent studies suggest that the normal colonisation of the newborn intestine is delayed and disturbed after birth by CS, which might disrupt the normal development of the immune system.28–31 Similarly, the hygiene hypothesis suggests that an overly clean environment and delayed or reduced exposure to infection in childhood may lead to a poorly regulated immune system and increase the risk of immune-mediated diseases. According to this hypothesis, children born by CS are less exposed to infections than children born vaginally.32 33 Furthermore, in vaginal delivery, the act of labour promotes the production of various cytokines activating the immune system of both the mother and child, which also differs in newborns delivered by CS.34 Alterations in the gut microbiota have been associated with the risk of MS in several studies15 and have also been found to influence the development and course of experimental autoimmune encephalomyelitis, an animal model for MS.35

Our study has several strengths. First, it includes a large population-based cohort followed up prospectively from fetal life onwards (MBRN), a nationwide registry for MS (MS Registry) and data on socioeconomic status (SN). The prospective nature of the study with big sample size and inherent statistical power excludes by design some of the biases innate to retrospective or cross-sectional studies. Our ability to combine data from various sources gives access to several relevant covariates, prevents loss to follow-up and ensures that only a small proportion of MS cases are missed. In addition, the MS cases in the MS Registry were verified by a neurologist according to the McDonald criteria, reducing the risk of misclassification.

Our study also has some limitations. Only individuals born in Norway were included in our study. So, our findings may not be generalisable to other populations. Further, we did not have information about maternal smoking during and before pregnancy for individuals born before 1999, so we could not evaluate possible confounding or effect modification by maternal smoking status. Nevertheless, a Swedish cohort study reported no association between maternal smoking during pregnancy and the risk of MS in their offspring.36 We did also not have information on the mother’s history of MS, which could have acted as a confounder as reports suggest that mothers with MS could be more likely to deliver by CS.37 However, we performed sibling-matched analyses to adjust for shared confounders by siblings, which would include the genetic predisposition inherited through the mother. Moreover, we did not have information on pediatric-onset of MS cases available for this study to assess whether delivery by CS is associated with early age at onset of MS. Nevertheless, MS is relatively rare in children,38 so we believe it would not affect the results by much. In addition, the incompleteness of the MS Registry might have influenced the results. However, the lack of registration was related to compliance with the registration process in neurological departments, which is unlikely related to the exposure or outcome of interest in this study. Therefore, a differential registration is unlikely, depending on the disease or characteristics of individuals with or without MS.

In summary, although our study suggests that there is a modest association between CS and the risk of MS, the stronger associations seen among individuals born preterm and individuals born with emergency CS who likely experienced a more complicated pregnancy/delivery may point to confounding underlying these associations. A greater understanding of fetal development and early maternal influences remains important in assessing risk factors for the development of MS.

Data availability statement

Data may be obtained from a third party and are not publicly available. Data may be obtained from a third party and are not publicly available. Data from the Norwegian Multiple Sclerosis Registry and MBRN are accessible to researchers by application.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by the regional committee for medical and health research ethics of Western Norway (reference/ID number REK 2017/298).Patients with multiple sclerosis (MS) gave written informed consent for their data to be used for research purposes when they were included in the Norwegian Multiple Sclerosis Registry and Biobank.

Acknowledgments

The authors acknowledge the Medical Birth Registry of Norway, the Norwegian Multiple Sclerosis Registry and Biobank and Statistics Norway for providing data for the study.

References

Footnotes

  • Contributors TR, MC and AKD conceptualised and designed the study. AKD and MC acted as guarantor for the study. AKD performed the acquisition of the data for the study. AK performed the analysis and interpretation of data for the work and drafting of the manuscript. AK, AKD, KB, EB, MP, TR, K-MM and MC interpreted the data and revised the work critically for important intellectual content. All authors gave the final approval of the version to be published. They agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy of integrity of any part of the work are appropriately investigated and resolved.

  • Funding This work was supported by Neuro-SysMed, Department of Neurology, Haukeland University Hospital, Bergen, Norway. Neuro-SysMed is funded by the Norges Forskningsråd (Norwegian Research Council, grant 288164).

  • Competing interests K-MM received speaker honoraria from Biogen, Novartis, Roche and Sanofi, and has participated in clinical trials organised by Biogen, Merck, Novartis, Roche and Sanofi.

  • Provenance and peer review Not commissioned; externally peer reviewed.