Incidence rates of cesarean deliveries vary substantially between countries, and account for approximately 19% of the deliveries worldwide.1 The main indications for cesarean deliveries are previous cesarean deliveries, suspected nonreassuring fetal status and labor dystocia.2 Rates of cesarean deliveries have increased in the last decades, and in some countries, rates reach >40% of deliveries.3 , 4 This inflation is mainly due to increasing rates of recurrent cesarean deliveries and an increase in other risk factors such as obesity.5 When a cesarean delivery takes place due to a medical indication, it is associated with lower mortality rates of the newborn, and it is considered a safe mode of delivery.6 However, cesarean deliveries have been linked to several short-term perinatal adverse outcomes including increased respiratory complications of the newborn.7–9
Recent studies have also suggested an association between mode of delivery and long-term health of the offspring. Obesity, impaired immune function, asthma, celiac, and inflammatory bowel disease were all found to occur at higher rates in cesarean-delivered offspring.10–15 The etiology of the association between mode of delivery and immune function may involve the different microbial exposures the newborn experiences during delivery.16 , 17 With increasing numbers of offspring delivered by cesarean deliveries, and as these offspring age, it is critical to clarify the potential long-term health implications of this trend.6 We aimed to study the association between delivery mode and the risk for pediatric long-term infectious morbidities of the offspring.
A population-based cohort analysis was performed, including all singleton infants born between the years 1991 and 2014 and discharged alive from Soroka University Medical Center (SUMC). SUMC includes the largest birth center in Israel, and it is the only tertiary hospital and pediatric ward in the Negev, Israel’s southern region.
The study population consisted of all singleton deliveries occurring during the study period. Multiple gestations, fetuses with congenital malformations and perinatal mortality cases (intrauterine, intrapartum or postpartum) were excluded (Figure 1). Study population was restricted to uncomplicated pregnancies and deliveries, which may compromise offspring health; therefore, all cases of nonreassuring fetal status, nonprogressive labor, cord prolapse, failed instrumental delivery, premature rupture of membranes, failed labor induction, placental abruption and pregnancy complications (including hypertensive disorders of pregnancy, placenta previa, gestational diabetes, Rh isoimmunization and maternal history of perinatal death) were excluded from the study. The remaining cesarean sections were considered elective, defined as cesarean delivery performed without initiation of labor, in a nonemergency setting. Most common indications include previous cesarean deliveries, malpresentation and/or maternal request. The study protocol was approved by the SUMC institutional review board (committee #0438-15-SOR), and informed consent was exempt.
Mode of delivery, which was the independent variable, was based on maternal delivery chart and was defined as vaginal or cesarean deliveries. The SUMC perinatal dataset consists of information recorded directly after delivery by an obstetrician. Skilled medical secretaries routinely review the information before entering it into the database. Coding is performed after assessment of medical prenatal care records together with routine hospital documents. All Israeli citizens are assigned a unique identifying number, either at birth or, if not born in Israel, upon receiving citizenship. This number is individual and in use throughout life, in all official offices and services. The perinatal and pediatric datasets were crosslinked based on maternal and offspring personal identifying numbers. Because SUMC is the single tertiary center in the region, with an emergency room, neonatal intensive care unit and pediatric departments, it is assumed that residents of the region will turn to SUMC when medical treatment is required.
The outcome variable (event) was defined as the first pediatric hospitalization of the offspring with any infectious morbidity diagnosis. The pediatric hospitalization dataset includes International Classification of Diseases, 9th revision (ICD-9) codes for all medical diagnoses, as well as demographic information.
The original SUMC pediatric dataset includes all diagnoses and was used to classify morbidities during hospitalizations of children up to 18 years of age, was classified into 19 categories of morbidities by a certified pediatrician (I.S.), by system and cause. These categories included infectious diseases–related diagnoses; cardiac, neurologic and other morbidities categories. All infectious morbidities were selected and classified into subcategories by origin and cause using the ICD-9 codes. For each child born at SUMC, all hospitalizations (visits to the emergency room that did not results in hospitalizations were not included) that included diagnoses of one of the infectious subcategories of morbidities were defined, and time to the first hospitalization was calculated from birth.
A list of the grouped infectious diagnoses and ICD-9 codes is presented in Supplementary Table 1, http://links.lww.com/INF/D384.
Follow-up time was calculated from birth to an event, death (if occurred in the hospital, and was not infectious disease related), 18 years of age or end of the study period (censored). If more than 1 infectious diagnosis was present at the first hospitalization, all were included in the univariable analysis comparing the rates based on study group.
Background characteristics included maternal age; birthweight (continuous); parity (1, 2–5, ≥5), gestational age (continuous); ethnicity (categorical variable: Jewish/Bedouin), gender, low Apgar scores 1 and 5 (all dichotomous), small for gestational age (defined as <5th percentile per gestational age and sex) and insufficient prenatal care (defined as ≤2 prenatal care visits and 1 documented prenatal testing results). Ethnicity is considered a proxy measure for socioeconomic status.
Statistical analysis was performed using STATA (STATA Crop. College Station, TX) and SPSS (IBM SPSS, Armonk, NY, IBM corp.) software. Assumptions were 2-sided with α = 0.05 and β = 0.2. Initial analysis compared background, pregnancy and perinatal characteristics between the study groups (ie, vaginal vs. cesarean deliveries), using χ2 and t test based on variable characteristics and normal distribution.
Gestational age estimation upon delivery was based on standard criteria using menstrual history and earliest ultrasound findings. Dating was based on last menstrual period (LMP), if it was certain and consistent with the ultrasound. In cases of a discrepancy between the ultrasound and LMP, or if the LMP was unknown, determination of gestational age was done according to first trimester sonographic dating. When no first trimester ultrasound was available, a second trimester ultrasound was used. Gestational age assessment by ultrasonography in the first part of the second trimester (between 14 0/7 and 21 6/7 weeks of gestation, inclusive) was based on a composite of fetal biometric measurements. If dating by ultrasonography performed between 14 0/7 weeks and 15 6/7 weeks of gestation (inclusive) varied from LMP dating by more than 7 days, or if ultrasonography dating between 16 0/7 weeks and 21 6/7 weeks of gestation varied by more than 10 days, the estimated due date was changed to correspond with the ultrasonography dating. If ultrasonography dating between 22 0/7 weeks and 27 6/7 weeks of gestation (inclusive) varied by more than 14 days from LMP dating, the estimated due date was changed to correspond with the ultrasonography dating. Date changes for smaller discrepancies were based on how early the ultrasound examination was performed and on clinician assessment of LMP reliability.
Incidence density rates per 1000 person-years were calculated and compared between the study groups. Kaplan–Meier survival curves were constructed, and the cumulative infectious-related hospitalization incidence was compared between the groups using the log rank test.
Variables statistically different between the study groups were tested for associations with the outcome variable. Variables associated with both the background characteristics and the outcome, and not known to be on the pathway between exposure and outcome, were suspected as confounders and considered in the multivariable model.18
Variables with clinical importance were also included in the multivariable analysis.
To adjust for length of follow-up, the multivariable Weibull parametric survival analysis was performed, in which mothers in the cohort were entered as clusters and the dependence among the siblings was accounted for. The models compared the independent risk for infectious disease hospitalization based on mode of delivery, and adjusted hazard ratios (adj. HRs) are presented. Separate models are presented for the total study population, and for term and preterm offspring separately, as offspring born preterms are at higher risk for morbidities. The final model was selected based on the lowest −2 log likelihood.
During the study period, 138,910 newborns met the inclusion criteria: 13,206 (9.5%) were delivered by cesarean deliveries, and 125,704 (91.5%) were delivered vaginally. Table 1 presents background information of the study population, including demographic, obstetrical and perinatal characteristics. When compared with the vaginally delivered group, mothers in the cesarean deliveries group were older, with higher parity and were less likely to present with insufficient prenatal care. Preterm births (<37 weeks gestation) and low birthweight (<2500 g) were also more common in the cesarean deliveries versus vaginal delivery group. Although low Apgar score 1 (<7) was more prevalent among the cesarean deliveries group, low Apgar score 5 (<7) was more prevalent among the vaginally delivered group.
During the follow-up period (0–18 years; median: 10.22), 13,054 children (9.4%) were hospitalized with infectious morbidity: 12.0% among the cesarean delivered children and 9.1% in the vaginally delivered group [Relative Risk 1.36; 95% confidence-interval (CI): 1.28–1.43]. The incidence density rates for first hospitalization were 15.22/1000 person-years and 9.06/1000 person-years for cesarean delivered and vaginally delivered, respectively (hazard ratio: 1.47; 95% CI: 1.39–1.55). Table 2 presents rates of selected diagnoses and total infection morbidities by the study groups, as well as time-adj. HRs. Cesarean-delivered offspring had higher rates of total infection-related hospitalizations, as well as meningitis, urinary tract infection, streptococci- or staphylococci-related infections, respiratory syncytial virus, otitis, upper respiratory tract infection and bronchiolitis. There were no differences between the groups in rates of hepatitis or fungal, Gram-negative, herpes, varicella and viral infections, or pneumonia (results not presented).
Figure, Supplemental Digital Content 1, http://links.lww.com/INF/D287, presents the Kaplan–Meier log of survival for the total infectious morbidities-related hospitalizations by study group. The overall difference between the groups was significant (Kaplan–Meier, log rank P < 0.001). When comparing total infectious morbidity–related hospitalizations between the groups up to the age of 1 month, rates were comparable between the groups (Kaplan–Meier, log rank P = 0.53); however, by the age of approximately 3 months old (91 days), the differences between the study groups became apparent and significant (Kaplan–Meier, log rank P = 0.023).
Table 3 presents results of the multivariable survival analysis models. The models compared the risk for total infectious morbidity–related hospitalizations among the study groups in total (model 1), and separately among term offspring (model 2) and preterm-delivered offspring (model 3). The association between cesarean deliveries and long-term pediatric infectious morbidity remained significant while controlling for maternal age, offspring gender, low Apgar (<7) at 5 minutes, insufficient prenatal care, birthweight or gestational age (adj. HR: 1.18, 95% CI: 1.11–1.25 and adj. HR: 1.10, 95% CI: 1.04–1.17 among the total study population, term only or preterm only, respectively).
In this large, population-based study with a long follow-up period, children delivered via elective cesarean deliveries were found to be at an increased risk for infectious morbidity–related pediatric hospitalizations, when compared with vaginally delivered children. This association was also observed among offspring delivered at term and preterm, separately, and was found to be independent of several important perinatal characteristics such as preterm births or gestational age, birthweight and low Apgar score.
Our study has several strengths, mainly being based on a single medical facility, serving all residents of the region, and with a long follow-up period on the offspring. Although loss of follow-up has probably occurred, there is no reason for it to be associated with the exposure and outcome and to therefore confound the associations found. Additionally, all hospitalization-related expenses are covered under National Health Insurance law, allowing all citizens access to obstetrical and pediatric care, and minimizing representation bias by sociodemographic/economic background. Other possible confounders, such as birthweight and gestational age, were available for analysis and were adjusted for in the multivariable analysis.
Several study limitations need to be addressed when interpreting our results. Our population was defined as uncomplicated pregnancies and deliveries, and the outcome assessed was infectious-related hospitalizations, thus is likely to represent acute and severe infectious diseases only. Therefore, the external validity of our findings is limited. Additionally, information regarding breast feeding history and antibiotic use after the delivery was unavailable. Both characteristics are related to delivery mode. Breast feeding is also known to promote beneficial bacteria, and may be confounding the association between mode of delivery and infectious morbidity of the offspring.10 , 19
Although newborns delivered vaginally are exposed to vaginal and perianal/gastrointestinal bacteria, a newborn delivered via cesarean deliveries is exposed to the operating room and nonmaternal dermal microbial colonization.16 , 17 Immunologic function and response differ based on the microbiotic colonizations, and because vaginal deliveries are associated with a more diverse neonatal microbiota, the developing immune system is exposed to a healthy diverse array of antigens. The differences in the microbiotica exposures of the newborn during delivery may alter immune reaction to infectious diseases, and may therefore account for infectious morbidities susceptibility. Animal studies have found different microbiome count and diversity between newborns based on mode of delivery, as early as age 1 month.20 The difference in flora based on mode of delivery may also be associated with the lower breast feeding rates after cesarean deliveries, as compared with vaginally delivered newborns as breast milk contains bacteria important for newborn flora and immune function.21
Although not directly addressed in the present study, our findings add support to several human studies suggesting that there is an association between mode of delivery and microbiota development, including the upper respiratory tract and gastrointestinal microbiota.22 Although the association between mode of delivery and gastrointestinal and respiratory tract infectious morbidities were expected, our results suggest additional organs and systems to be involved with increased susceptibility for infectious morbidity, including meningitis, otitis and urinary tract infections.
The increased overall infectious morbidity suggest a more general immune disruption, which is in agreement with the hygiene hypothesis, which proposes that suboptimal bacterial exposure in early life is associated with a greater risk of developing immune diseases later in life.23
Several mechanisms have been suggested to improve offspring immune system function and reduce morbidities susceptibility, including maternal pre- and postnatal nutritional strategies and probiotics supplementation for the offspring.24 , 25
Further studies, preferably with a prospective design, may help to clarify the relationship between mode of delivery and overall as well as specific infectious morbidity. This will allow factors such as breast feeding, antibiotic use and other maternal and pediatric characteristics that may be possible confounders to be addressed.
Our findings add an important element to be considered when counseling women regarding delivery mode.
1. Betran AP, Ye J, Moller AB, et al. The increasing trend in caesarean section rates: global, regional and national estimates: 1990–2014. PLoS One. 2016;11:e0148343.
2. Timofeev J, Reddy UM, Huang CC, et al. Obstetric complications, neonatal morbidity, and indications for cesarean delivery by maternal age. Obstet Gynecol. 2013;122:1184–1195.
3. Althabe F, Sosa C, Belizán JM, et al. Cesarean section rates and maternal and neonatal mortality in low-, medium-, and high-income countries: an ecological study. Birth. 2006;33:270–277.
4. Mi J, Liu F. Rate of caesarean section is alarming in China. Lancet. 2014;383:1463–1464.
5. Menacker F, Declercq E, Macdorman MF. Cesarean delivery: background, trends, and epidemiology. Semin Perinatol. 2006;30:235–241.
7. Levine EM, Ghai V, Barton JJ, et al. Mode of delivery and risk of respiratory diseases in newborns. Obstet Gynecol. 2001;97:439–442.
8. Gerten KA, Coonrod DV, Bay RC, et al. Cesarean delivery and respiratory distress syndrome: does labor make a difference? Am J Obstet Gynecol. 2005;193(3 pt 2):1061–1064.
9. Hammad IA, Chauhan SP, Magann EF, et al. Peripartum complications with cesarean delivery: a review of Maternal-Fetal Medicine Units Network publications. J Matern Fetal Neonatal Med. 2014;27:463–474.
10. Cho CE, Norman M. Cesarean section and development of the immune system in the offspring. Am J Obstet Gynecol. 2013;208:249–254.
11. Yuan C, Gaskins AJ, Blaine AI, et al. Association between cesarean birth and risk of obesity in offspring in childhood, adolescence, and early adulthood. JAMA Pediatr. 2016;170:e162385.
12. Black M, Bhattacharya S, Philip S, et al. Planned repeat cesarean section at term and adverse childhood health outcomes: a record-linkage study. PLoS Med. 2016;13:e1001973.
13. Bager P, Simonsen J, Ethelberg S, et al. Cesarean delivery and risk of intestinal bacterial infection. J Infect Dis. 2010;201:898–902.
14. Bager P, Simonsen J, Nielsen NM, et al. Cesarean section and offspring’s risk of inflammatory bowel disease: a national cohort study. Inflamm Bowel Dis. 2012;18:857–862.
15. Kristensen K, Henriksen L. Cesarean section and disease associated with immune function. J Allergy Clin Immunol. 2016;137:587–590.
16. Dominguez-Bello MG, Costello EK, Contreras M, et al. Delivery mode
shapes the acquisition and structure of the initial microbiota across multiple body habitats in newborns. Proc Natl Acad Sci U S A. 2010;107:11971–11975.
17. Biasucci G, Benenati B, Morelli L, et al. Cesarean delivery may affect the early biodiversity of intestinal bacteria. J Nutr. 2008;138:1796S–1800S.
18. Ananth CV, Schisterman EF. Confounding, causality, and confusion: the role of intermediate variables in interpreting observational studies in obstetrics. Am J Obstet Gynecol. 2017;217:167–175.
19. Siggers RH, Thymann T, Jensen BB, et al. Elective cesarean delivery affects gut maturation and delays microbial colonization but does not increase necrotizing enterocolitis in preterm pigs. Am J Physiol Regul Integr Comp Physiol. 2008;294:R929–R938.
20. Harmsen HJ, Wildeboer-Veloo AC, Raangs GC, et al. Analysis of intestinal flora development in breast-fed and formula-fed infants by using molecular identification and detection methods. J Pediatr Gastroenterol Nutr. 2000;30:61–67.
21. Bosch AATM, Levin E, van Houten MA, et al. Development of upper respiratory tract microbiota in infancy is affected by mode of delivery. EBioMedicine. 2016;9:336–345.
22. Strachan DP. Hay fever, hygiene, and household size. BMJ. 1989;299:1259–1260.
23. Nauta AJ, Ben Amor K, Knol J, et al. Relevance of pre- and postnatal nutrition to development and interplay between the microbiota and metabolic and immune systems. Am J Clin Nutr. 2013;98:586S–593S.
24. West CE, Hammarström ML, Hernell O. Probiotics during weaning reduce the incidence of eczema. Pediatr Allergy Immunol. 2009;20:430–437.
25. Biesbroek G, Bosch AA, Wang X, et al. The impact of breastfeeding on nasopharyngeal microbial communities in infants. Am J Respir Crit Care Med. 2014;190:298–308.