Perinatal and Other Risk Factors for Common Infections in Infancy: A Prospective Cohort Study

Objective: Limited data from prospective cohort studies in high-income countries are available on the perinatal risk factors for common infections in children. Our hypothesis was that perinatal factors may be risk factors for infectious episodes during the first year of life. Methods: In this prospective Health and Early Life Microbiota birth cohort study of full-term infants (n = 1052) born in 2016–2018, the number and duration of infection episodes were collected online at weekly to monthly intervals. In a multivariate regression model, the main exposures were perinatal factors such as mode of delivery and intrapartum antibiotics. Environmental factors were additional exposures. The outcomes were the number and duration of infectious episodes in the first year of life. Results: The mean number of infection episodes was 4.2 (2.9 SD). The mean duration of infection symptoms was 44 days (40 SD). Upper respiratory infections accounted for 83% of the episodes (3674/4455). Perinatal factors were not associated with the number nor the duration of infection episodes, but cesarean section was associated with an increased occurrence of urinary tract infections in infancy [adjusted odds ratio (aOR): 3.6; 95% confidence interval (CI): 1.13–11.1]. Of the additional exposures male sex (aOR: 1.1; 95% CI: 1.0–1.2) and the presence of siblings (aOR: 1.3; 95% CI: 1.2–1.4) were associated with the number of infection episodes. Conclusions: This prospective cohort study showed that perinatal factors, mode of delivery and intrapartum antibiotics were not associated with the risk of common infections in infancy, but cesarean delivery was associated with a risk of urinary tract infections.

groups did not show any significant differences in baseline characteristics.
We have reported baseline characteristics of the HELMi cohort earlier. 22Briefly, the proportion of infants born via cesarean section was 17% (Table 1), reflecting the general proportion of cesarean section births in Finland (16.7% in 2017). 23In addition to previously reported characteristics, the children were exclusively breastfed up to a mean age of 3.3 months (1.8 SD) and partially breastfed up to 10 months (3.3 SD).By 12 months of age 77% of the children were fully vaccinated against rotavirus, pneumococcus, diphtheria-tetanus-pertussis-polio-Hib, measles-mumps-rubella and varicella-zoster virus in accordance with the national immunization program.Altogether 4.8% of the children started day care in the first year of life.In addition to the regular electronic online data collection from the families, 97% (n = 1015/1052) of the infants attended a health status examination at the age of 3 months. 22he study was conducted in accordance with the principles of the Helsinki Declaration and the protocol was reviewed by the ethical committee of the Helsinki and Uusimaa Hospital District (263/13/03/03 2015), Finland.All the families gave their written informed consent in advance.

Clinical Follow-up
A prospective clinical follow-up was conducted using electronic questionnaires that stored the answers instantly in an online database.The parents filled in the questionnaire form weekly for the first 4 months, every 2 weeks until 7 months and monthly from 8 to 12 months.In addition, they filled in comprehensive questionnaires that included childcare practices, living conditions and infant health at 3-month intervals.The parents received automatic reminders via email and text message to promote compliance with this questionnaire program. 22amilies recorded daily infection episodes and infection-related symptoms and reported them in a continuous manner using an online study diary.The infections recorded in this way were classified as follows: (1) Respiratory tract infections (RTI): upper (URTI) and lower (LRTI) respiratory tract infections, (2) URTI: common cold, cough, laryngitis, influenza, otitis media, tonsillitis and pharyngitis, (3) LRTI: pneumonia, acute obstructive bronchitis, bronchiolitis, (4) GI: diarrhea, vomiting, (5) UTI: pyelonephritis and any UTI and (6) other infections (see

Statistical Analysis
The main exposures of interest were delivery mode and exposure to intrapartum antibiotics.Other exposures and covariates were the number of siblings, breastfeeding, the use of probiotics during pregnancy, parental asthma, maternal education status, day-care attendance and presence of furry pets in the household.The outcomes were the number of infection episodes and the number of infection-related symptom days during the first year of life.For the statistical analysis, an infection episode was defined by the first and last dates with symptoms that the families reported.If the dates of different symptoms overlapped or started on the day after another symptom, they were considered to form 1 episode.The infections were categorized according to the symptoms.
We used a multivariate Poisson regression model to estimate the association between exposures and outcomes.Means, SDs, medians and interquartile ranges (IQR) were reported for the infections.The multivariate analyses included the following variables: sex, year and season of birth, mode of delivery, intrapartum antibiotics, number of siblings, parental asthma, maternal education, use of probiotics during pregnancy and presence of furry pets in the household.Maternal education and the presence of siblings were grouped into 2 category variables in the analyses.The analyses were performed with SPSS version 28, and the results were expressed as adjusted odds ratios with 95% confidence intervals (95% CIs).We estimated the required sample size for the observational cohort study using StatsDirect Statistical Software.We regarded OR of 1.25 as a clinically significant risk for both cesarean section and exposure to intrapartum antibiotics during vaginal birth.We chose statistical power of 80% and alpha error of 5%.The required sample size, with a ratio of 1:4 for exposed and unexposed subjects, was 165 exposed and 660 unexposed participants.

Infections in the Cohort
The 1052 full-term infants who met the inclusion criteria (Fig. 1) and were followed up with online data collection for the first year of life had a total of 4455 infection episodes, affecting 90% of them (Table 2).The mean numbers of infections per child for the year and for each 3-month period can be seen in Table 2.
Antibiotics were prescribed for 9.2% of the infection episodes, including 10% of URTIs, 71% of the LRTIs, 92% of the UTIs, 0.8% of the GIs and 4.5% of the other infections.Amoxicillin was the most frequently used antimicrobial agent (57%), followed by amoxicillin-clavulanate (18%).The proportion of all infections that were treated in the hospital was 2.4%.Severe infections such as osteomyelitis, cellulitis and sepsis were rare (0.25%) and none of the children died.

Perinatal Factors
Multivariate analysis showed no association between the number of overall infection episodes and perinatal factors such as mode of delivery and intrapartum antibiotics (Table 3).The differences in the infection burden between cesarean and vaginal delivery were not statistically significant (95% CI: 0.85-1.01)(Table 3).Children exposed to intrapartum antibiotics had a smaller infection burden (95% CI: 0.88-1.03)than those not exposed to antibiotics, which again were not statistically significant findings (Table 3).

Perinatal Factors
UTI was the only infection with a statistically significant association with the mode of delivery, cesarean delivery being associated with an increased mean number of UTIs (95% CI: 1.1-11.1) in the first year of life as compared with those born via vaginal delivery (Table 3).The difference in days of reported UTI-related symptoms was not statistically significant (Table 3).There was no difference in the number of UTI episodes between the infants exposed to intrapartum antibiotics and those not exposed (Table 3).

Other Factors
The use of probiotics during pregnancy was a protective factor for UTIs, as it was associated with a decreased mean number of UTI episodes (95% CI: 0.1-0.99) in the first year as compared with children not exposed to probiotics during pregnancy.The consumption of probiotics during pregnancy was not statistically associated with the number of reported UTI symptom days (Table 4).

DISCUSSION
In this prospective birth cohort study of 1052 infants, perinatal factors, including delivery mode and exposure to intrapartum antibiotics, were not associated with the number of infection episodes during the first year of life or the total number of days with infection symptoms, although the risk of UTIs was associated with the delivery mode, as the infants born via cesarean delivery had an increased risk.

Perinatal Factors and Infections
To our knowledge, there are a limited number of prospective birth cohort studies performed in high-income countries that concern perinatal factors, previously shown to modify the colonization and development of the gut and respiratory microbiome, 24,25 as risk factors for common infections in childhood.Two previous Scandinavian birth cohort studies addressing the effect of cesarean delivery on the risk of childhood infections have reported inconsistent results, 4,26 while in Denmark, Vissing et al. 4 found that cesarean delivery increased the risk of LRTIs.A large birth cohort study conducted in Norway was nevertheless unable to find any significant association between cesarean delivery and recurrent LRTIs. 26Here, we found no association between perinatal factors (mode of delivery and intrapartum antibiotics) and the occurrence of overall infections or RTIs.
Other factors associated with common infections in childhood have been investigated more extensively.In 5 prospective birth cohort studies from Denmark, 4,10 Germany 1,3 and Holland 2 day-care attendance and the presence of siblings have been reported as risk factors for infections in childhood.The occurrence of infection episodes in the first year of life obtained from our cohort was comparable to those reported in 2 of the above-mentioned birth cohorts studied in Denmark and Germany, 1,4 whereas the numbers of infections reported in the other cohorts were higher than in the present study. 2,3,10However, the differences in the methods of reporting symptoms need to be noted when comparing these results.Only one of the previous birth cohort studies made partial use of electronic data collection, 3 whereas the present data were collected prospectively with an electronic online system.UTI has been reported to be one of the leading reasons for intravenous antibiotic treatment for hospitalized children in a high-income country with a comprehensive immunization program. 27UTIs have previously been associated with the composition of the gut microbiota, 18 which in turn has also been associated with perinatal factors such as mode of delivery 24 and intrapartum antibiotics. 25We found no association between intrapartum antibiotics and UTIs, but we did find cesarean delivery to be associated with an increased occurrence of UTIs.This observation supports the idea that there is a link between the maturing gut microbiota and the risk of UTI.

Hyvönen et al
The results of this observational cohort study suggest that the use of probiotics during pregnancy might reduce the occurrence of UTIs in infants.Because of the small sample of UTIs in this cohort, we were not able to assess the effects of different probiotics, the 3 most used being Lactobacillus rhamnosus, Lactobacillus acidophilus and Bifidobacterium bifidum.In any case, the results need to be interpreted with caution because of the high proportion of mothers using probiotics during pregnancy (71%).In particular, the evidence of vertical transfer of prenatal probiotics is inconclusive. 28hen assessing other associated factors, our results complement previous findings, that male sex is a risk factor for RTIs 2,8,29 and that the existence of older siblings is negatively associated with RTIs. 2,4,10,121][32] The only protective factor we found was exposure to a furry pet, which was positively associated with the number of overall infections, but not with RTIs as has previously been established. 1,33,34he strength of our study lies in the large prospective longitudinal cohort providing comprehensive prospective data on 1052 families.The extensive questionnaires and diaries were collected online at weekly to monthly intervals, and there were few cases of missing data, as most families continued the follow-up until the child reached 1 year of age.As the main focus of the HELMi-birth cohort is gut microbiota development, 20,35 the data includes over 10,000 biological samples in total (fecal, breastmilk and genetic testing), which can be connected to our study's infection data in future studies.The strength of the present epidemiological cohort study is that the association of the gut microbiota composition with infectious diseases in infancy in this prospective cohort is underway.This enables the comparison of epidemiological perinatal risk factors presented here and those retrieved from a large-scale bioinformatics analysis regarding gut microbiota composition, and their associations with the occurrence of infectious diseases in infants.We acknowledge that the sample size may be a limitation for some included comparisons.The main limitation of this work is that the cohort represents highly educated mothers in a high-income country.Almost all the mothers breastfed their infants, for instance and very few reported smoking.Thus, the results are not generalizable to child cohorts with different background characteristics.As the number of children starting day care in the first year of life was low, we were not able to assess the impact of day care on the infection burden at this age.It is to be noted that our study shows only associations rather than causality.

CONCLUSIONS
The results of this prospective cohort study show that perinatal factors, which have previously been shown to be related to the development of the human microbiota, were not associated with the common infections found in infants, except for an increased occurrence of UTIs in those born via cesarean section.The present findings suggest that the burden of common respiratory infections in infants does not depend on the mode of delivery or intrapartum antibiotics, but is more closely related to other environmental factors.

TABLE 2 .
Incidence and Duration of Infectious Episodes in 1052 Full-term Children During Their First Year of Life

TABLE 3 .
Mode of Delivery and Intrapartum Antibiotics as a Risk Factor for the Occurrence of Infections in Fullterm Infants During Their First Year of Life Values are aORs and 95% CIs obtained by using a Poisson regression model.*All the aORs were adjusted for sex, year of birth, mode of delivery, intrapartum antibiotics, number of siblings, parental asthma, maternal education, use of probiotics during pregnancy and a furry pet in the household.aORindicates adjusted odds ratio; CI, confidence interval.

TABLE 4 .
Other Risk Factors for the Occurrence of Infections in Full-term Infants During Their First Year of Life All the aORs were adjusted for sex, year of birth, mode of delivery, intrapartum antibiotics, number of siblings, parental asthma, maternal education, use of probiotics during pregnancy and a furry pet in the household.†Limitednumber of observations.-indicates not applicable; aOR, adjusted odds ratio; CI, confidence interval.