Characteristics of the study populations are shown in Table 1. In MoBa, the prevalence of asthma at 7 years was 4.1%. In the register-cohort, 3.8% of the children had asthma at 7 years, while 4.4% had asthma at 13 years of age. Overall, 15.6% of the children in MoBa were exposed to antibiotics during pregnancy, while 28.3% of children were exposed to antibiotics during pregnancy in the register-based cohort (Supplementary material Tables S4 and S6). Extended spectrum penicillins, such as amoxicillin were the most commonly used antibiotics in both study populations (Supplementary material Table S3).
Maternal antibiotic use during pregnancy and offspring asthma
Any maternal antibiotic use during pregnancy was associated with asthma at 7 year in both study populations (aRR 1.23, 95% CI 1.11-1.37 in MoBa and aRR 1.21, 95% CI 1.16 - 1.25 in register cohort) after adjusting for maternal characteristics (Fig. 1 and 2). In MoBa, the estimate decreased but remained significant after additional adjustment for RTIs and UTIs during pregnancy (aRR 1.15, 95% CI 1.02-1.30). We observed a dose-response pattern between antibiotics during pregnancy and offspring asthma in MoBa, where the use of antibiotics for two or more infections yielded an aRR of 1.54 (95% CI 1.27-1.88) after adjusting for infections. However, no dose-response association was observed in the register-based cohort.
In MoBa, exposure to antibiotics only in the late pregnancy (≥ 17 weeks) was associated with greater risk of asthma than use only in the early pregnancy (<17 weeks), when comparing these mutually exclusive time windows to no use of antibiotics during pregnancy (Fig. 1). In the register-based cohort, antibiotic use during early and late pregnancy was similar (Fig. 2). After sibling-matched analysis of antibiotic use in the register cohort, associations disappeared (Supplementary material Table S8).
A weak association with any antibiotic prescriptions to mothers during pregnancy was also found with the risk of offspring asthma at 13 years (aRR 1.13, 95% CI 1.03-1.23) in the register-based cohort (Fig. 2). Similar to the association with asthma at 7 years, we did not observe dose-response pattern or time-dependent relationship in the association between antibiotic exposure during pregnancy and risk of asthma at 13 years.
Adjustment for child characteristics and exposures (mediators) further decreased the estimates (Supplementary material Table S4 for MoBa and Tables S6 and S7 for register-based cohort).
Maternal infections during pregnancy and offspring asthma in MoBa
Children of mothers with any LRTIs (aRR 1.30, 95% CI 1.07-1.57), two or more URTIs (aRR 1.19, 95% CI 1.09 - 1.30), and any UTIs (aRR 1.26, 95% CI 1.11-1.42) during pregnancy, had an increased risk of asthma at 7 years (Fig. 1). URTIs and any UTIs showed consistent associations with child asthma after additional adjustment for antibiotic use during pregnancy (Fig. 1). Adjustment for offspring characteristics and exposures (potential mediators) slightly decreased the estimates (Supplementary material Table S5).
Maternal antibiotic use conditional on specific infection and offspring asthma in MoBa
We observed that maternal antibiotic use during pregnancy was not associated with increased risk of offspring asthma beyond what was seen for those exposed to a particular indication such as LRTI and UTI (Fig. 3). However, there was a significantly higher risk associated with antibiotics use conditional on URTI. Supplementary material Table S9 shows that children of mothers with LRTI during pregnancy who did not use antibiotics had a higher risk of asthma (aRR 1.40, 95% CI 1.09-1.79) than children of mothers who used antibiotics for LRTI (aRR 1.26, 95% CI 0.93-1.71), when compared to children of mothers without any use of antibiotics or LRTIs. Similar result was observed related to UTIs during pregnancy (Supplementary material Table S9).
Antibiotic use outside of pregnancy, paternal antibiotic use and offspring asthma
In the register-based cohort, maternal antibiotic use before and after pregnancy showed similar associations with the risk of asthma as use during pregnancy (Supplementary material Table S6 and S7). In MoBa, maternal use of antibiotics during pregnancy showed higher risk of child asthma than the use after pregnancy (Supplementary material Table S4), but information on antibiotic use before pregnancy was not available. Father’s use of antibiotics during pregnancy was not associated with offspring asthma at 7 years after adjustment for paternal characteristics (aRR 1.04, 95% CI 0.67-1.61) in MoBa.
Prescription indication and offspring asthma in the register cohort
We observed increased risk of asthma at 7 years for antibiotics most likely prescribed for RTIs (aRR 1.18, 95% CI 1.13-1.24) and UTIs (aRR 1.16, 95% CI 1.11-1.22) after adjustment for confounders (Supplementary material Table S10). Association was observed also between prescription for RTIs (aRR 1.14, 95% CI 1.02-1.26) and asthma at 13 years.