Abstract
Background:
Previous studies have linked prenatal acetaminophen use to increased asthma risk in children. However, none have explored this association while differentiating between asthma cases with and without other allergic conditions or by employing objective biomarkers to assess acetaminophen exposure.
Objective:
Examine whether the detection of acetaminophen biomarkers in cord blood is associated with the subgroups of asthma both with and without allergic comorbidities in children.
Methods:
Acetaminophen biomarkers, including unchanged acetaminophen and acetaminophen glucuronide, were measured in neonatal cord blood samples from the Boston Birth Cohort. Asthma subgroups were defined based on physician diagnoses of asthma and other allergic conditions (atopic dermatitis, allergic rhinitis). Multinomial regressions were used to examine the associations between acetaminophen biomarkers and asthma subgroups, adjusting for multiple confounders, including potential indications for maternal acetaminophen use such as maternal fever.
Results:
The study included 142 children with asthma and at least one other allergic condition, 55 children with asthma but no other allergic condition, and 613 children free of asthma. Detection of acetaminophen in cord blood, reflecting maternal exposure to acetaminophen shortly before delivery, was associated with a 3.73 times the odds of developing asthma without allergic comorbidities (95% CI: 1.79, 7.80, p=0.0004). In contrast, detection of acetaminophen in cord blood was not associated with elevated risk of asthma with allergic comorbidities. Analysis of acetaminophen glucuronide yielded consistent results.
Conclusion:
In a prospective birth cohort, cord blood acetaminophen biomarkers were associated with an increased risk of childhood asthma without allergic comorbidities, but were not associated with childhood asthma with allergic comorbidities.
Keywords: childhood asthma, maternal acetaminophen use, cord blood metabolites, allergic conditions
Introduction
Asthma is the most common chronic disease in children globally, with a prevalence of 10%−15% in Europe and North America from 2000–20171. Approximately 80% of children with asthma also experience other allergic comorbidities, such as allergic rhinitis and atopic dermatitis2. Among this subgroup, allergic asthma were more common, while children with non-allergic asthma rarely have allergic conditions3. Studies have shown that allergic asthma and non-allergic asthma differ in molecular and celluar characteristics4 and have different environmental and genetic risk factors5,6. Therefore, studies examining the association between early life exposures and subsequent asthma risk should distinguish between asthma cases with and without allergic comorbidities.
Acetaminophen is the most commonly used over the counter (OTC) medication to relieve mild to moderate pain and treat fever in pregnant woman. It is present in over 600 OTC and prescription drugs as an active ingredient7. Evidence suggests that acetaminophen can cross the placenta and alter fetal development7,8. Large-scale observational studies have reported associations between prenatal acetaminophen exposure and increased risk of childhood asthma9–13. However, these previous findings have limitations that warrant addressing for a robust evaluation of the relationship between prenatal acetaminophen exposure and childhood asthma risk. Firstly, existing studies did not differentiate between asthma cases with and without allergic comorbidities. Secondly, prenatal acetaminophen exposure was determined through self-report or prescription electronic records, which are prone to errors from inaccurate recall and potential oversight of acetaminophen exposures in various OTC drugs. Finally, few studies have explored the modification effect by low birthweight, a condition that may affect a newborn’s acetaminophen metabolism14 and consequently the effects on asthma outcomes later in life.
We aimed to address these limitations and investigate the prospective associations between objective biomarkers of acetaminophen exposure in cord blood and later school-age asthma, considering cases with and without other allergic conditions. We analyzed data from the Boston Birth Cohort (BBC), a long-standing prospective US birth cohort. Our study sample comprised 810 children in the BBC with measured acetaminophen biomarkers in cord plasma, reflecting maternal acetaminophen use shortly before delivery. All study participants have been followed for at least 6 years since birth, and their asthma and other relevant clinical information were determined based on physician diagnoses in the electronic medical records (EMRs).
Methods
Study design and population
We utilized data from the Boston Birth Cohort (BBC) study for the current analysis. The BBC represents a predominantly urban, low-income, and underserved population in the Boston area, Massachusetts, United States. Detailed descriptions of the BBC study design have been published previously15,16. In brief, mothers who delivered a singleton live birth at the Boston Medical Center (BMC) were enrolled at delivery. A subset of children receiving pediatric care at the BMC were prospectively followed until 21 years of age. A random sample of children under follow-up (n=1000) was selected to participate in a metabolomic sub-study. Metabolome profiling, including acetaminophen metabolites, was performed using umbilical cord plasma samples, as detailed below. Participants who were followed up until at least 6 years old and had available measurements of acetaminophen metabolites were eligible for the current analysis. Eligible participants were enrolled between December 2002 and October 2013.
The BBC study protocol was approved by the Institutional Review Boards of the BMC and the Johns Hopkins Bloomberg School of Public Health. Written consent was obtained from all participating mothers.
Definition of asthma subgroups
We defined three mutually exclusive subgroups based on physician diagnoses of asthma, atopic dermatitis (AD), and allergic rhinitis (AR) recorded in the EMR using International Classification of Diseases codes (eMethods). The subgroups are asthma with allergic comorbidities, asthma without allergic comorbidities, and no asthma. The asthma with any allergic comorbidity group includes children with an asthma diagnosis at 6-year-old or older and any diagnosis of AD or AR from birth to the end of follow-up. The asthma without allergic comorbidities group includes children with an asthma diagnosis at 6-year-old or older and no AD or AR from birth to the end of follow-up. The no asthma group includes children who were free of asthma diagnosis at 6-year-old or older, regardless of their diagnosis of AD and AR.
Acetaminophen biomarkers in cord plasma
Umbilical cord blood samples were collected at birth. Quantitative profiling of metabolites, including acetaminophen, acetaminophen glucuronide, and 3-(N-acetyl-L-cystein-S-yl)-acetaminophen in cord plasma was performed at the Harvard-MIT Broad Institute Metabolite Profiling Laboratory using liquid chromatography-tandem mass spectrometry (LC-MS), following the established protocol described in a previous publication17. Multiple quality control and quality assurance procedures have been conducted following the established protocol17. Briefly, a pooled reference sample composed of all individual study samples was randomly inserted across samples (per 20–30 samples) and coefficient of variation (CV) was calculated for each metabolite using the reference samples18. The CV for acetaminophen, acetaminophen glucuronide, and 3-(N-acetyl-L-cystein-S-yl)-acetaminophen was 0.07, 0.06, and 0.03, respectively, suggesting high reliability of the measurements.
We closely inspected the raw intensity output from LC-MS and determined the background noise level for each acetaminophen metabolites. Samples with a specific metabolite value above the corresponding background noise level were determined to have a detectable value for that metabolite. We created a dichotomized variable (detected versus not detected) based on the background noise level for each acetaminophen metabolite.
Due to the skewed distributions of the raw intensities, we conducted rank-based inverse normal transformation (INT). Before transformation, non-detectable values were imputed as one half of the minimal detectable value of the specific metabolite. After the normalization, the effects of potential outliers were controlled.
Details of measurement, quality control, and normalization of the acetaminophen metabolite data are described in the eMethods.
Potential confounders
We collected child and maternal characteristics through maternal interviews at the time of study enrollment within 24–72 hours of delivery and/or EMR abstraction. Child covariates were sex, preterm birth (gestational age <37 weeks), low birthweight (birth weight <2500g), parity, and delivery type. Maternal covariates were age at delivery, education, marital status, race and ethnicity, any cigarette smoking during pregnancy, any gestational conditions (gestational diabetes, preeclampsia, or HELLP syndrome), intrauterine inflammation (any placenta histopathology consistent with uterine inflammation that was determined by pathologists or the presence of intrapartum maternal fever >38°C at parturition), pre-pregnancy body mass index (BMI), history of asthma, and stress level during pregnancy.
Statistical analysis
We used multinomial regression models to examine the associations between asthma subgroups and the detection of each cord acetaminophen metabolite, adjusting for the potential confounders as covariates. Individuals with no missing value in the covariates were included in the regression models. As sensitivity analyses, we fitted the multinomial regression models using inverse normally transformed metabolite values (continuous) as an independent variable. Additionally, we compared the group of participants with detectable values for all the three acetaminophen biomarkers versus those without detectable value for any of the three acetaminophen biomarkers. To explore whether low birthweight may modify the associations between acetaminophen biomarkers and the asthma subgroups, we stratified the study sample low birth weight (birth weight <2500g) and fitted adjusted multinomial regression models in each stratum. All statistical analyses were performed using R version 4.1.1.
Results
Among a total of 810 eligible participants, 142 (17.5%) had asthma and at least one allergic condition (allergic rhinitis or atopic dermatitis), 55 (6.8%) had asthma but no allergic condition, and 613 (75.7%) were free of asthma. The mean age at the last study visit was 10.7 years and was not significantly different across the asthma subgroups. eTable 1 shows the distribution of total serum immunoglobulin E (IgE) concentration during childhood (age range: 1 to 8.5 years) among a subset of participants (n=167, 20.6%) with the available measurement. We observed a higher total IgE level among children with asthma and allergic comorbidities (median [IQR]: 69.93 [28.55–420.33]; mean: 406.03 kU/L) compared to children with asthma but no allergic comorbidities (median [IQR]: 45.88 [11.63–122.12]; mean: 98.62 kU/L); however, the difference was not statistically significant (Wilcoxon rank-sum p-value: 0.13).
Overall, 59% of the participants have a non-Hispanic Black mother and 23% have a Hispanic mother. As shown in Table 1, regardless of allergic comorbidities, participants with asthma were more likely to have male sex, preterm birth, low birth weight, mother not married, prenatal smoking, higher maternal pre-pregnancy BMI, and maternal history of asthma, compared to participants without asthma. Additionally, comparing to children with asthma and allergic comorbidities, children who had asthma without allergic comorbidities were more likely to be female, have low birth weight, maternal gestational conditions, higher maternal pre-pregnancy BMI, and were less likely to have intrauterine inflammation.
Table 1.
Maternal and child characteristics by asthma subgroups at age of 6 years (n=810), Boston Birth Cohort
| Variable | Asthma with allergic comorbidities | Asthma without allergic comorbidities | No asthma | P-value |
|---|---|---|---|---|
| Number of participants | 142 (17.5) | 55 (6.8) | 613 (75.7) | |
| Age in years at the last visit, mean (SD) | 11.04 (3.10) | 10.90 (2.88) | 10.66 (3.09) | 0.377 |
| Male | 99 (69.7) | 32 (58.2) | 325 (53.0) | 0.001 |
| Maternal race and ethnicity | 0.293 | |||
| Non-Hispanic Black | 96 (67.6) | 30 (54.5) | 351 (57.3) | |
| Non-Hispanic White | 6 (4.2) | 3 (5.5) | 29 (4.7) | |
| Hispanic | 28 (19.7) | 16 (29.1) | 145 (23.7) | |
| Others | 12 (8.5) | 6 (10.9) | 88 (14.4) | |
| Preterm birth (gestational age <37 weeks) | 38 (26.8) | 15 (27.3) | 91 (14.8) | 0.001 |
| Low birth weight (birth weight <2500 grams) | 29 (20.4) | 18 (32.7) | 92 (15.0) | 0.002 |
| Parity: primiparous | 63 (44.4) | 19 (34.5) | 242 (39.5) | 0.391 |
| Cesarean section | 48 (33.8) | 25 (45.5) | 205 (33.6) | 0.202 |
| Maternal education | 0.904 | |||
| High school or less | 98 (70.0) | 36 (67.9) | 416 (68.1) | |
| Some college | 28 (20.0) | 12 (22.6) | 120 (19.6) | |
| College degree or above | 14 (10.0) | 5 (9.4) | 75 (12.3) | |
| Maternal marital status: not married | 107 (76.4) | 40 (72.7) | 377 (62.3) | 0.003 |
| Maternal age in years at delivery, mean (SD) | 28.09 (7.12) | 28.22 (6.63) | 28.73 (6.68) | 0.549 |
| Any maternal smoking during pregnancy | 41 (29.1) | 15 (27.3) | 89 (14.7) | <0.001 |
| Any maternal gestational condition | 17 (12.0) | 15 (27.8) | 89 (14.6) | 0.018 |
| Gestational diabetes mellitus | 8 (5.6) | 8 (14.8) | 42 (6.9) | |
| Preeclampsia | 9 (6.3) | 10 (18.2) | 51 (8.3) | |
| HELLP syndrome | 2 (1.4) | 0 (0.0) | 3 (0.5) | |
| Intrauterine inflammation | 26 (18.8) | 5 (9.4) | 59 (9.9) | 0.011 |
| Maternal pre-pregnancy BMI, mean (SD) | 27.07 (6.38) | 29.33 (9.13) | 26.46 (5.96) | 0.008 |
| Maternal history of asthma | 37 (28.5) | 19 (35.8) | 63 (11.2) | <0.001 |
| Maternal stress level over lifetime | 0.188 | |||
| Not stressful | 58 (40.8) | 16 (29.1) | 253 (41.5) | |
| Average | 70 (49.3) | 29 (52.7) | 301 (49.3) | |
| Very Stressful | 14 (9.9) | 10 (18.2) | 56 (9.2) | |
| Maternal stress level during pregnancy | 0.123 | |||
| Not stressful | 45 (31.9) | 18 (32.7) | 243 (39.9) | |
| Average | 66 (46.8) | 23 (41.8) | 272 (44.7) | |
| Very Stressful | 30 (21.3) | 14 (25.5) | 94 (15.4) |
Abbreviations: BMI, body mass index; HELLP, hemolysis, elevated liver enzymes, and low platelet count; SD, standard deviation. Data are presented as number (percentage) of individuals unless otherwise indicated. P-values were obtained using Pearson χ2 tests (or Fisher’s exact test for small cells) for categorical variables and analysis of variance tests for continuous variables.
Acetaminophen, acetaminophen glucuronide, and 3-(N-acetyl-L-cysteine-S-yl)-acetaminophen were detected in 16.4%, 17.5%, and 28.4% of the cord plasma samples, respectively. The distributions of the LC-MS raw intensities are shown in eFigure 1. There were high correlations between each pair of the acetaminophen biomarkers. Among samples with detectable acetaminophen, 94% had detectable 3-(N-acetyl-L-cysteine-S-yl)-acetaminophen. All samples with detectable acetaminophen glucuronide had detectable 3-(N-acetyl-L-cysteine-S-yl)-acetaminophen (eTable 2). Additionally, in 121 samples, all three biomarkers were detected and in 572 samples, none of the biomarkers were detected.
Participants with detectable cord acetaminophen were more likely to have preterm birth, primiparous parity, cesarean section, prenatal smoking, gestational conditions, intrauterine inflammation, maternal history of asthma, and maternal stress than participants without detectable acetaminophen (Table 2).
Table 2.
Maternal and child characteristics by detection of acetaminophen in cord blood (n=810), Boston Birth Cohort
| Acetaminophen detection by LC-MS | |||
|---|---|---|---|
| Variable | Not detected | Detected* | P-value |
| Number of participants | 677 (83.6) | 133 (16.4) | |
| Male | 374 (55.2) | 82 (61.7) | 0.205 |
| Maternal race and ethnicity | 0.185 | ||
| Non-Hispanic Black | 404 (59.7) | 73 (54.9) | |
| Non-Hispanic White | 27 (4.0) | 11 (8.3) | |
| Hispanic | 157 (23.2) | 32 (24.1) | |
| Others | 89 (13.1) | 17 (12.8) | |
| Preterm birth (gestational age <37 weeks) | 105 (15.5) | 39 (29.3) | <0.001 |
| Low birth weight (birth weight <2500 grams) | 109 (16.1) | 30 (22.6) | 0.093 |
| Parity: primiparous | 260 (38.4) | 64 (48.1) | 0.046 |
| Cesarean section | 219 (32.4) | 59 (44.4) | 0.011 |
| Maternal education | 0.965 | ||
| High school or less | 461 (68.6) | 89 (67.4) | |
| Some college | 133 (19.8) | 27 (20.5) | |
| College degree or above | 78 (11.6) | 16 (12.1) | |
| Maternal marital status: not married | 429 (64.3) | 95 (71.4) | 0.14 |
| Maternal age in years at delivery, mean (SD) | 28.67 (6.67) | 28.13 (7.18) | 0.401 |
| Any maternal smoking during pregnancy | 112 (16.8) | 33 (24.8) | 0.038 |
| Any maternal gestational condition | 83 (12.4) | 38 (28.6) | <0.001 |
| Gestational diabetes mellitus | 48 (7.1) | 10 (7.5) | |
| Preeclampsia | 39 (5.8) | 31 (23.3) | |
| HELLP syndrome | 3 (0.4) | 2 (1.5) | |
| Intrauterine inflammation | 64 (9.7) | 26 (20.3) | 0.001 |
| Maternal history of asthma | 86 (13.6) | 33 (28.7) | <0.001 |
| Maternal pre-pregnancy BMI, mean (SD) | 26.74 (6.38) | 26.80 (5.95) | 0.921 |
| Maternal stress level over lifetime | 0.004 | ||
| Not stressful | 283 (42.0) | 44 (33.1) | |
| Average | 334 (49.6) | 66 (49.6) | |
| Very Stressful | 57 (8.5) | 23 (17.3) | |
| Maternal stress level during pregnancy | 0.044 | ||
| Not stressful | 266 (39.5) | 40 (30.3) | |
| Average | 300 (44.6) | 61 (46.2) | |
| Very Stressful | 107 (15.9) | 31 (23.5) | |
Abbreviations: BMI, body mass index; HELLP, hemolysis, elevated liver enzymes, and low platelet count; LC-MS, liquid chromatography-tandem mass spectrometry; SD, standard deviation. Data are presented as number (percentage) of individuals unless otherwise indicated. P-values were obtained using Pearson χ2 tests (or Fisher’s exact tests for small cells) for categorical variables and Student’s t-test for continuous variables.
Detected is defined as levels above the background noise.
Figure 1 shows the distributions of cord acetaminophen metabolites (INT values) by asthma subgroups. We observed more individuals with a higher value (right tail of the distribution) for the three acetaminophen biomarkers in the group of children with asthma but no allergic comorbidities, comparing to the other two asthma subgroups. Acetaminophen and acetaminophen glucuronide were detected in 36.4% and 30.9% of participants who had asthma but no allergic comorbidities, respectively, which were statistically significantly higher than the other asthma subgroups (Table 3). The proportions of individuals with detection of acetaminophen metabolites among children with asthma and allergic comorbidities did not seem to differ from those among children free of asthma (Table 3).
Figure 1. Distributions of cord acetaminophen biomarkers by childhood asthma subgroups.
The horizontal axis shows the ranked inverse normal transformed intensity of (a) acetaminophen, (b) acetaminophen glucuronide, and (c) 3-(N-acetyl-L-cysteine-S-yl) acetaminophen in cord plasma. The vertical axis shows the density of the value for each biomarker.
Table 3.
Descriptive statistics of cord blood acetaminophen biomarkers by asthma subgroups at age of 6 years (n=810)
| Cord plasma acetaminophen biomarkers | Asthma with allergic comorbidities | Asthma without allergic comorbidities | No asthma | P-value |
|---|---|---|---|---|
| Number of participants | 142 (17.5) | 55 (6.8) | 613 (75.7) | |
| Acetaminophen | ||||
| Ranked inverse transformed intensity, mean (SD) | 0.07 (0.64) | 0.42 (0.90) | 0.04 (0.63) | <0.001 |
| Detection of the biomarker by LC-MS | <0.001 | |||
| Not detected | 118 (83.1%) | 35 (63.6%) | 524 (85.5%) | |
| Detected* | 24 (16.9%) | 20 (36.4%) | 89 (14.5%) | |
| Acetaminophen glucuronide | ||||
| Ranked inverse transformed intensity, mean (SD) | 0.09 (0.69) | 0.34 (0.89) | 0.05 (0.64) | 0.008 |
| Detection of the biomarker by LC-MS | 0.024 | |||
| Not detected | 117 (82.4%) | 38 (69.1%) | 513 (83.7%) | |
| Detected* | 25 (17.6%) | 17 (30.9%) | 100 (16.3%) | |
| 3-(N-acetyl-L-cysteine-S-yl)-acetaminophen | ||||
| Ranked inverse transformed intensity, mean (SD) | 0.10 (0.78) | 0.25 (0.89) | 0.06 (0.73) | 0.178 |
| Detection of the biomarker by LC-MS | 0.557 | |||
| Not detected | 101 (71.1%) | 36 (65.5%) | 443 (72.3%) | |
| Detected* | 41 (28.9%) | 19 (34.5%) | 170 (27.7%) | |
Abbreviations: LC-MS, liquid chromatography-tandem mass spectrometry; SD, standard deviation. Data are presented as number (percentage) of individuals unless otherwise indicated. P-values were obtained using Pearson χ2 tests for categorical variables and analysis of variance tests for continuous variables.
Detected is defined as levels above the background noise.
Figure 2 shows the associations between cord acetaminophen biomarkers and asthma subgroups. After adjusting for potential confounders, participants with detectable acetaminophen had 3.73 times the odds of developing asthma without allergic comorbidities compared to participants without detectable acetaminophen (95% CI: 1.79, 7.80, p=0.0004). Consistently, individuals with detectable acetaminophen glucuronide were also more likely to develop asthma without allergic comorbidities than those without detectable values (OR=2.16, 95% CI: 1.02, 4.58, p=0.0439). Detection of 3-(N-acetyl-L-cysteine-S-yl)-acetaminophen was not significantly associated with increased risk of asthma without allergic comorbidities (OR=1.25, 95% CI: 0.61, 2.56, p=0.5399). However, among a subset of participants with detectable values for all three acetaminophen metabolites, we observed a significantly increased risk of asthma without allergic comorbidities compared to participants with no acetaminophen metabolites detected in cord blood (OR=2.51, 95% CI: 1.12, 5.61, p=0.0108). We did not find significant associations between any of the cord acetaminophen biomarkers and asthma with allergic comorbidities in the overall sample after adjusting for covariates (e.g., OR=0.69, 95% CI: 0.36, 1.32, p=0.2637 for acetaminophen). Adjusting for gestational age and birth weight in a continuous fashion did not change the results. The findings were consistent when the INT metabolite values (continuous) were used as independent variables in the multinomial regression models (eTable 3).
Figure 2. Associations between cord acetaminophen biomarkers and childhood asthma subgroups.
The odds ratios (ORs) and 95% confidence intervals (95% CIs) were estimated from multinomial regressions adjusting for sex, preterm birth, low birth weight, primiparous parity, delivery type, maternal race and ethnicity, maternal age, maternal education, maternal marital status, prenatal smoking, maternal pre-pregnancy BMI, any gestational conditions, intrauterine inflammation, maternal asthma, maternal stress during pregnancy.
To minimize potential confounding by indication, we conducted sensitivity analyses excluding participants with maternal conditions that may lead to maternal acetaminophen use, including intrauterine inflammation, intrapartum maternal fever, preeclampsia, and maternal asthma. The associations between acetaminophen biomarker at birth and childhood asthma we observed in the main analysis remained unchanged in these sensitivity analyses (eTable 4).
In the stratified analysis (Table 4), a higher cord acetaminophen level was significantly associated with an increased risk of asthma without allergic comorbidities in both birth weight strata, with a larger effect size in the birth weight <2500g stratum (OR=4.43 per SD increase in INT-acetaminophen, 95% CI: 1.39, 14.14, p=0.0120) than in the birth weight ≥2500g stratum (OR=1.81, 95% CI: 1.09, 3.02, p=0.0224). Acetaminophen level was not significantly associated with asthma with allergic comorbidities in either the birth weight <2500g stratum (OR=1.34, 95% CI: 0.58, 3.10, p=0.4911) or the birth weight ≥2500g stratum (OR=0.66, 95% CI: 0.41, 1.07, p=0.0895). The other two metabolites showed the same direction of effects for asthma subgroups when stratifying by low birth weight. Additional subgroup analysis stratified by preterm birth yielded consistent results (eTable 5).
Table 4.
Adjusted associations between cord blood acetaminophen biomarkers (inverse-normal transformed; continuous) and asthma subgroups, stratified by low birth weight
| No asthma | Asthma without allergic comorbidities | Asthma with allergic comorbidities | ||
|---|---|---|---|---|
| Acetaminophen | ||||
| Birth weight ≥2500g | N | 521 | 37 | 113 |
| OR | 1 (Ref.) | 1.81 | 0.66 | |
| 95% CI | - | (1.09,3.02) | (0.41,1.07) | |
| P-value | - | 0.0224 | 0.0895 | |
| Birth weight <2500g | N | 92 | 18 | 29 |
| OR | 1 (Ref.) | 4.43 | 1.34 | |
| 95% CI | - | (1.39,14.14) | (0.58,3.10) | |
| P-value | - | 0.0120 | 0.4911 | |
| Acetaminophen glucuronide | ||||
| Birth weight ≥2500g | N | 521 | 37 | 113 |
| OR | 1 (Ref.) | 1.32 | 0.65 | |
| 95% CI | - | (0.77,2.28) | (0.41,1.03) | |
| P-value | - | 0.3115 | 0.0663 | |
| Birth weight <2500g | N | 92 | 18 | 29 |
| OR | 1 (Ref.) | 2.84 | 1.38 | |
| 95% CI | - | (1.13,7.14) | (0.67,2.84) | |
| P-value | - | 0.0268 | 0.3828 | |
| 3-(N-Acetyl-L-cystein-S-yl)-acetaminophen | ||||
| Birth weight ≥2500g | N | 521 | 37 | 113 |
| OR | 1 (Ref.) | 1.11 | 0.71 | |
| 95% CI | - | (0.65,1.92) | (0.48,1.04) | |
| P-value | - | 0.6970 | 0.0757 | |
| Birth weight <2500g | N | 92 | 18 | 29 |
| OR | 1 (Ref.) | 2.07 | 1.73 | |
| 95% CI | - | (0.83,5.12) | (0.85,3.52) | |
| P-value | - | 0.1167 | 0.1316 | |
Abbreviations: N, number of participants; OR, odds ratio; 95% CI, 95% confidence interval. ORs were estimated from multinomial regression models in which each continuous acetaminophen biomarker (ranked inverse transformed) was an independent variable. Low birth weight was defined as birth weight <2500g. The adjusted model includes the following covariates: sex, gestational age at birth, primiparous parity, delivery type, maternal race and ethnicity, maternal age, maternal education, maternal marital status, prenatal smoking, maternal pre-pregnancy BMI, any gestational conditions, intrauterine inflammation, maternal asthma, maternal stress during pregnancy.
Discussion
In a sample of 810 children representing an urban, low-income, underserved population from Boston, detection of acetaminophen and acetaminophen glucuronide in cord blood was associated with an increased risk of developing asthma without allergic comorbidities at 6-year-old or older. The positive association of cord acetaminophen level with the subgroup of asthma without allergic comorbidities was stronger in children with low birth weight. We did not find significant association between acetaminophen biomarkers and the subgroup of asthma with allergic comorbidities in this study population. Given the inherent difficulty in reliably diagnosing asthma in younger children, we focused our examination on participants who had reached at least 6 years of age by the time of the last study visit.
The main strength of this study was the use of objective measures of acetaminophen at birth. These objective cord biomarkers are less prone to measurement errors due to inaccurate self-report or prescription records that do not capture the use of a large number of OTC medication containing acetaminophen. Pharmacokinetic studies of acetaminophen have reported that in pregnant women during the third trimester, oral administration of acetaminophen can reach maximum plasma concentration within an hour19. The half-life of a therapeutic dose of acetaminophen in late pregnancy was estimated to be 1.4–3.7 hours19–22, and the pharmacokinetic parameters are comparable between fetal cord blood and maternal venous blood in full-term fetuses with normal birth weights22. Therefore, the metabolites in this study likely reflect maternal use of acetaminophen a few hours before delivery. In our study, we detected acetaminophen, acetaminophen glucuronide, and 3-(N-acetyl-L-cysteine-S-yl)-acetaminophen in 16.4%, 17.5%, and 28.4% of the cord plasma samples, respectively. In 14.9% of the samples, all three metabolites were detected. The level of 3-(N-acetyl-L-cysteine-S-yl)-acetaminophen was highest in the subgroup of samples with all three biomarkers detected, followed by samples with two and one biomarkers detected (eTable 6). The larger proportion of detectable 3-(N-acetyl-L-cysteine-S-yl)-acetaminophen than the other two biomarkers may be due to differences in clearance time, other biological factors involved in acetaminophen metabolism, and/or assay specificity.
Concerns have been raised in studies that the observed association between prenatal acetaminophen use and childhood asthma may be largely explained by unmeasured confounding23,24. In this study, we adjusted for a relatively comprehensive list of potential confounders, including many maternal characteristics not controlled in previous studies. These include intrauterine inflammation (including maternal intrapartum fever), maternal gestational conditions, maternal stress level, and maternal pre-pregnancy BMI. These maternal clinical factors can cause symptoms such as fever and mild to moderate pain, likely increasing the use of over-the-counter (OTC) or prescription medications containing acetaminophen. To further address this concern, we conducted additional sensitivity analyses and showed that among individuals who were free of maternal indications for acetaminophen use, the association between acetaminophen biomarker at birth and childhood asthma without allergic comorbidities persisted. These results suggest that the associations we observed in the main analysis were robust across subgroups of participants with different maternal indications for use of acetaminophen.
Given the widespread use of acetaminophen in pregnant people, our study focused on assessing the associations between acetaminophen biomarkers measured in their offspring at birth and later asthma risk of the children. While it is possible that postnatal exposures, such as respiratory tract infection and acetaminophen use during early childhood, may also influence asthma risk in children, this data was not available for the current analysis, which has limited our ability to explore the interplay between acetaminophen levels at birth and postnatal risk factors. Nevertheless, it is noteworthy that, since the acetaminophen biomarkers were measured at birth in our study, the measurements were not subject to the influence of postnatal exposures. Future studies examining the combined effect of maternal acetaminophen use near delivery and early childhood risk factors would be valuable to elucidate the complex dynamics involved in asthma development. In addition, other postnatal data, such as lung functions measured by spirometry, will also be valuable to confirm physician’s diagnosis of asthma from electronic medical records in pediatric population.
Acetaminophen toxicity is mainly introduced by N-Acetyl-p-benzoquinone imine (NAPQI), a toxic intermediate formed during the phase I oxidation of acetaminophen metabolism25 (eFigure 2). NAPQI is normally detoxified by glutathione (GSH) and forms nontoxic 3-(N-acetyl-L-cysteine-S-yl)-acetaminophen25. GSH plays important roles in counteracting the oxidative stress produced by reactive oxygen species. As the level of acetaminophen exposure increases, continued production of NAPQI will eventually deplete GSH, weakening the ability of the host to defend against oxidative stress in airway13,26,27. Therefore, depletion of GSH has been proposed as the biologic mechanism that links acetaminophen exposure to asthma13. A recent animal study provided further evidence that developing lungs are susceptible to acetaminophen toxicity through increased pulmonary expression of CYP2E1, the enzyme that catalyzes the phase I oxidation of acetaminophen metabolism28.
We observed a more pronounced positive association between cord acetaminophen levels and the subgroup of asthma without allergic comorbidities in children with low birth weight than in those with normal birth weight. This suggests that children with low birth weight are more susceptible to the effects of acetaminophen exposure around delivery. Low birth weight is a recognized risk factor for childhood asthma29,30, thought to primarily due to smaller lungs and hyperreactive airways rather than allergic sensitization31. Additionally, neonates with low birth weight exhibit reduced liver function32 and hepatic drug metabolizing capacity33,34. Consequently, exposure to acetaminophen in this subgroup is likely prolonged, potentially explaining the larger effect size for developing asthma without allergic comorbidities.
To our knowledge, only one small retrospective study has investigated the effects of in-utero acetaminophen exposure on asthma subtype with a focus on allergic asthma. This study compared nineteen children from Singapore who had allergic asthma to their non-allergic, non-asthmatic siblings and found that the case group had more mother-reported prenatal acetaminophen use than the controls35. Large prospective studies from the United Kingdom found that acetaminophen use during late pregnancy was associated with increased risks of wheezing and total immunoglobulin E (IgE) level in early to middle childhood but the exposure was not associated with allergic rhinitis, atopic dermatitis, or skin test positivity36,37. In our sample, we did not observe a statistically significant association between the acetaminophen biomarkers in cord blood and the subgroup of asthma with allergic comorbidities.
Acetaminophen is considered the preferred option to relieve pain and fever in pregnant women38 since non-steroidal anti-inflammatory drugs (NSAIDs) are contraindicated during pregnancy39. Ongoing debates surround public health recommendations for acetaminophen use in pregnant women7,40. While evidence has suggested prenatal acetaminophen exposure may be linked to an increased risk of neurodevelopmental and reproductive disorders in children7, very few, if any, medication options exist for pregnant women to relieve pain and fever. In this analysis, we observed diverse associations between asthma with and without allergic comorbidities and between children with low and normal birth weight. Our findings suggest that research on acetaminophen use during pregnancy should carefully evaluate potentially heterogeneous effects on child health outcomes in various subgroups and tailor recommendations based on a careful risk-benefit calculation considering both effect size and the prevalence of disease subtypes. Additionally, it is critical that recommendations appropriately consider the needs of pregnant women with health conditions requiring medication treatment and offer relief options.
While our study has several strengths, several points need consideration regarding our results. First, the cord acetaminophen biomarkers were measured only once, and the exact dose, time, and duration of exposure were unknown. Given the short half-life of acetaminophen, these biomarkers can only capture acetaminophen exposure shortly before delivery. Future studies should investigate whether the findings can be generalized to exposure at earlier times during pregnancy. Second, our metabolome panel did not capture other major acetaminophen metabolites (e.g., acetaminophen sulfate, NAPQI), limiting our ability to quantify the overall burden of acetaminophen exposure. Third, although we tried our best to account for potential confounders, residual confounding by unmeasured or imperfectly measured variables is still possible in this observational study. Future work with study designs that may better address this challenge is needed. Fourth, our asthma subgroup classifications were based on medical record reports of allergic conditions. While total serum IgE levels available in a subset of our study participants were appropriately correlated with the subgroup classifications, i.e. higher IgE levels among participants classified as having asthma with allergic co-morbidities, it is possible there was some subgroup misclassification. Future studies that collect IgE levels through skin prick or serologic testing in all study participants at the time of asthma diagnosis may offer better subgroup classification. Lastly, the sample size, especially for the subgroup of asthma without allergic comorbidities, was modest. Future studies should replicate our results in larger cohorts and prioritize this asthma subgroup in the study design.
In conclusion, this is the first prospective cohort study to report an elevated risk of developing asthma without allergic comorbidities during school age in children with detected acetaminophen biomarkers in cord blood. This association may be more pronounced in children with low birth weight. We did not find a significant association between cord acetaminophen biomarkers and the subtype of asthma with allergic comorbidities. It is worth highlighting that the current study is observational by study design, thus the association reported in this study is not sufficient on its own to conclude acetaminophen biomarkers, measured at birth, cause asthma without co-occurring allergic conditions. Future investigations with a larger sample size and different study designs, e.g. clinical trials, are warranted to replicate our results and to build causal evidence. These findings may provide data on acetaminophen relationships with asthma that is important for clinicians, parents, and regulatory agencies to consider when evaluating approaches to prevent asthma in the population. If future research suggests a causal relationship, the impact of acetaminophen on asthma risk in the population as a whole need to be considered since non-allergic asthma occurs in a relatively small proportion of the population. Additionally, the impact of acetaminophen needs to be evaluated in the context of other known risk and protective factors for asthma and needs to consider the benefits of pregnant people taking acetaminophen to treat heath conditions.
Funding Source:
The Boston Birth Cohort (the parent study) was supported in part by the National Institutes of Health (NIH) grants (2R01HD041702, R21HD066471, R01HD086013, R01HD098232, R21AI154233, R01ES031272, R01ES031521, and U01ES034983) and by the Health Resources and Services Administration (HRSA) of the U.S. Department of Health and Human Services (HHS) (UT7MC45949). This information or content and conclusions are those of the author and should not be construed as the official position or policy of, nor should any endorsements be inferred by the funding agencies.
Dr. Ladd-Acosta reports receiving consulting fees from the University of Iowa for providing expertise on epigenetics outside of this work.
Abbreviations/Acronyms:
- AD
atopic dermatitis
- AR
allergic rhinitis
- BBC
Boston Birth Cohort
- BMC
Boston Medical Center
- BMI
body mass index
- CI
confidence interval
- CV
coefficient of variation
- CYP2E1
Cytochrome P450 Family 2 Subfamily E Member 1
- EMR
electronic medical records
- GSH
glutathione
- HELLP
hemolysis, elevated liver enzymes, low platelet count
- IgE
immunoglobulin E
- INT
inverse normal transformation
- LC-MS
liquid chromatography-tandem mass spectrometry
- NAPQI
N-Acetyl-p-benzoquinone imine
- NSAIDs
non-steroidal anti-inflammatory drugs
- OR
odds ratio
- OTC
over the counter
- SD
standard deviation
eMethods
ICD codes used to identify diagnosis in EMR
Since January 2004, the primary and secondary diagnosis for all visits (outpatient, emergency department, inpatient) at the Boston Medical Center were recorded in the electronic medical record using ICD-9 or ICD-10 codes. We identified cases for asthma (ICD-9-CM codes: 493.0–493.9; ICD-10-CM codes: J45.2-J45.5, J45.9, J82.83), atopic dermatitis (ICD-9-CM codes: 691.8; ICD-10-CM codes: L20), and allergic rhinitis (ICD-9-CM codes: 477.0, 477.2, 477.8, 477.9; ICD-10-CM codes: J30.9, J30.89, J30.81, J30.1, J30.2) based on physician diagnosis.
Metabolites profiling and quality control/assurance procedures
Metabolite profiling, including acetaminophen metabolites, was conducted using the hydrophilic interaction liquid chromatography in the positive ionization mode (HILIC-pos) analyses. The experimental design including sample preparation, HILIC-pos parameters, and quality control/assurance procedures, has been described in details in previous publications1,2. Specifically, the quality control and quality assurance procedures for metabolomic data are as follows. First, specimens were randomly located and were analyzed in a blinded fashion. Second, a pooled study reference sample composed of all study samples was randomly inserted across samples (per 20–30 samples); the coefficient of variation (CV) for each metabolite was calculated using the reference samples, and metabolites with CV > 20% were excluded from the downstream analyses. Third, 51 pairs of blind duplicate pooled samples were included to ensure data accuracy and reliability; these samples demonstrated extremely high levels of pairwise correlation (median: 0.9995, interquartile range: 0.9990 to 0.9998). Fourth, as internal standard metabolites, the same concentrations of 2 metabolites (phenylalanine- d8 and valine-d8) were respectively included in all study samples for HILIC-pos analyses; the CVs of these internal standard metabolites were <5%. Fifth, internal standard peak was monitored to ensure system performance throughout the analyses. Lastly, we closely inspected the raw intensities from the HILIC-pos to determine the background noise level for each of the three acetaminophen metabolites. The background noise levels are 43000, 597, and 700 for acetaminophen, acetaminophen glucuronide, and 3-(N-acetyl-L-cystein-S-yl) acetaminophen, respectively (eFigure 1). Samples with a specific metabolite value above the corresponding background noise level were determined to have a detectable value for that metabolite.
Ranked-based inverse normal transformation of metabolites
We applied ranked-based inverse normal transformation to the raw intensities of the metabolites from the HILIC-pos analyses using the following R code: qnorm((rank(x,na.last=“keep”)-0.5)/sum(!is.na(x))). After the normalization, the effects of potential outliers were controlled. Normalization also made the effect sizes across metabolites comparable when the continuous biomarker levels were used in statistical analysis.
The ranked-based inverse normal transformation replaces the sample quantiles by quantiles from the standard normal distribution without changing the sample ranks. Therefore, when grouping the samples based on their quantiles, the grouping will be the same regardless of whether the raw intensity or the inverse normal transformed intensity was used.
Descriptive statistics of participants characteristics
The maternal and child characteristics were compared across the diagnosis groups and between acetaminophen detection categories (detected versus not detected or noise), respectively, using Pearson χ2 tests (or Fisher’s exact tests if numbers are small) for categorical variables and analysis of variance (ANOVA) tests (or Student’s t-test) for continuous variables. The detection and the distribution of cord acetaminophen biomarkers were also compared across the diagnosis groups using Pearson χ2 tests and ANOVA tests, respectively.
eTable 1.
Distribution of total serum immunoglobulin E (IgE) concentration during childhood by asthma subgroups.
| No asthma (n=613) | ||||||||
|---|---|---|---|---|---|---|---|---|
| Min. | 1st quartile | Median | Mean | 3rd quartile | Max. | N (%) participants with available data | Mean age at measurement (years) | |
| Total IgE concentration (kU/L) | 0 | 7.83 | 18.02 | 78.61 | 55.53 | 2368.83 | 129 (21.0%) | 2.7 |
| Asthma with allergic comorbidities (n=142)† | ||||||||
| Min. | 1st quartile | Median | Mean | 3rd quartile | Max. | N (%) participants with available data | Mean age at measurement (years) | |
| Total IgE concentration (kU/L) | 2.74 | 28.55 | 69.93 | 406.03 | 420.33 | 2603.72 | 28 (19.7%) | 4.7 |
| Asthma without allergic comorbidities (n=55)‡ | ||||||||
| Min. | 1st quartile | Median | Mean | 3rd quartile | Max. | N (%) participants with available data | Mean age at measurement (years) | |
| Total IgE concentration (kU/L) | 2.36 | 11.63 | 45.88 | 98.62 | 122.12 | 329.952 | 10 (18.2%) | 3.6 |
The total IgE level among participants with asthma and allergic comorbidities were statistically significantly higher than that in the no asthma group (Wilcoxon rank-sum p-value: 3.4×10−5).
The total IgE level among participants with asthma but no allergic comorbidities was not statistically significantly different from that in the no asthma group (Wilcoxon rank-sum p-value: 0.28), or from that in the asthma with allergic comorbidities group (Wilcoxon rank-sum p-value: 0.13).
eTable 2.
Cross-tabulation of acetaminophen biomarkers detection
| Acetaminophen | |||
|---|---|---|---|
| Not detected or noise | Detected (above noise level) | ||
| 3-(N-Acetyl-L-cystein-S-yl)-acetaminophen | Not detected or noise | 572 | 8 |
| Detected (above noise level) | 105 | 125 | |
| Acetaminophen glucuronide | |||
| Not detected or noise | Detected (above noise level) | ||
| 3-(N-Acetyl-L-cystein-S-yl)-acetaminophen | Not detected or noise | 580 | 0 |
| Detected (above noise level) | 88 | 142 | |
| Acetaminophen | |||
| Not detected or noise | Detected (above noise level) | ||
| Acetaminophen glucuronide | Not detected or noise | 656 | 12 |
| Detected (above noise level) | 21 | 121 | |
eTable 3.
Associations between cord blood acetaminophen biomarkers (inverse-normal transformed; continuous) and asthma subgroups at age of 6 years or older
| No asthma (n=511) | Asthma without allergic comorbidities (n=43) | Asthma with allergic comorbidities (n=116) | ||
|---|---|---|---|---|
| Acetaminophen | ||||
| Unadjusted | OR | 1 (Ref.) | 1.86 | 1.07 |
| 95% CI | - | (1.35,2.56) | (0.81,1.42) | |
| P-value | - | 0.0001 | 0.6236 | |
| Adjusted | OR | 1 (Ref.) | 2.10 | 0.78 |
| 95% CI | - | (1.38,3.19) | (0.53,1.13) | |
| P-value | - | 0.0005 | 0.1817 | |
| Acetaminophen glucuronide | ||||
| Unadjusted | OR | 1 (Ref.) | 1.66 | 1.09 |
| 95% CI | - | (1.19,2.3) | (0.83,1.43) | |
| P-value | - | 0.0025 | 0.5227 | |
| Adjusted | OR | 1 (Ref.) | 1.54 | 0.80 |
| 95% CI | - | (1.02,2.31) | (0.56,1.14) | |
| P-value | - | 0.0379 | 0.2174 | |
| 3-(N-Acetyl-L-cystein-S-yl)-acetaminophen | ||||
| Unadjusted | OR | 1 (Ref.) | 1.35 | 1.07 |
| 95% CI | - | (0.98,1.88) | (0.85,1.37) | |
| P-value | - | 0.0695 | 0.5569 | |
| Adjusted | OR | 1 (Ref.) | 1.26 | 0.88 |
| 95% CI | - | (0.84,1.9) | (0.65,1.19) | |
| P-value | - | 0.2697 | 0.3990 | |
Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval. ORs were estimated from multinomial regression models in which each continuous acetaminophen biomarker (ranked inverse transformed) was an independent variable. The adjusted model includes the following covariates: sex, preterm birth, low birth weight, primiparous parity, delivery type, maternal race and ethnicity, maternal age, maternal education, maternal marital status, prenatal smoking, maternal pre-pregnancy BMI, any gestational conditions, intrauterine inflammation, maternal asthma, maternal stress during pregnancy.
eTable 4.
Association between cord acetaminophen biomarker and childhood asthma subgroup among all participants and participants without potential maternal indications for acetaminophen use.
| All participants (main analysis) | ||||
|---|---|---|---|---|
| No Asthma | Asthma with allergic comorbidities | Asthma without allergic comorbidities | ||
| Number of participants | 511 | 116 | 43 | |
| Acetaminophen (detected vs. not detected) | Adjusted OR (95% CI) | 1 (reference) | 0.69 (0.36, 1.32) | 3.73 (1.79, 7.80) |
| p-value | n/a | 0.2637 | 0.0004 | |
| Excluding participants with intrauterine inflammation or maternal fever | ||||
| No Asthma | Asthma with allergic comorbidities | Asthma without allergic comorbidities | ||
| Number of participants | 459 | 94 | 41 | |
| Acetaminophen (detected vs. not detected) | Adjusted OR (95% CI) | 1 (reference) | 0.81 (0.39, 1.71) | 3.72 (1.73, 7.99) |
| p-value | n/a | 0.5874 | 0.0008 | |
| Excluding participants with maternal preeclampsia | ||||
| No Asthma | Asthma with allergic comorbidities | Asthma without allergic comorbidities | ||
| Number of participants | 472 | 108 | 34 | |
| Acetaminophen (detected vs. not detected) | Adjusted OR (95% CI) | 1 (reference) | 0.61 (0.29, 1.27) | 4.79 (2.10, 10.90) |
| p-value | n/a | 0.1879 | 0.0002 | |
| Excluding participants with maternal asthma | ||||
| No Asthma | Asthma with allergic comorbidities | Asthma without allergic comorbidities | ||
| Number of participants | 453 | 81 | 29 | |
| Acetaminophen (detected vs. not detected) | Adjusted OR (95% CI) | 1 (reference) | 0.87 (0.39, 1.95) | 4.80 (1.94, 11.87) |
| p-value | n/a | 0.7354 | 0.0007 | |
The odds ratios (ORs) and 95% confidence intervals (95% CIs) were estimated from multinomial regressions. All analyses were adjusted for sex, preterm birth, low birth weight, primiparous parity, delivery type, maternal race and ethnicity, maternal age, maternal education, maternal marital status, prenatal smoking, maternal pre-pregnancy BMI, and maternal stress during pregnancy. Any gestational conditions, intrauterine inflammation, and maternal asthma were also adjusted for as covariates in the analyses where participants were not excluded based on these variables.
eTable 5.
Adjusted associations between cord blood acetaminophen biomarkers (inverse-normal transformed; continuous) and asthma subgroups, stratified by preterm birth
| No asthma | Asthma without allergic comorbidities (n=43) | Asthma with allergic comorbidities (n=116) | ||
|---|---|---|---|---|
| Acetaminophen | ||||
| Gestational age ≥37wks | N | 522 | 40 | 104 |
| OR | 1 (Ref.) | 2.19 | 0.63 | |
| 95% CI | - | (1.29,3.72) | (0.38,1.05) | |
| P-value | - | 0.0038 | 0.0774 | |
| Gestational age <37wks | N | 91 | 15 | 38 |
| OR | 1 (Ref.) | 2.94 | 1.21 | |
| 95% CI | - | (1.13,7.65) | (0.59,2.47) | |
| P-value | - | 0.0268 | 0.5951 | |
| Acetaminophen glucuronide | ||||
| Gestational age ≥37wks | N | 522 | 40 | 104 |
| OR | 1 (Ref.) | 1.57 | 0.65 | |
| 95% CI | - | (0.93,2.64) | (0.4,1.05) | |
| P-value | - | 0.0939 | 0.0797 | |
| Gestational age <37wks | N | 91 | 15 | 38 |
| OR | 1 (Ref.) | 2.27 | 1.36 | |
| 95% CI | - | (1.01,5.12) | (0.74,2.51) | |
| P-value | - | 0.0476 | 0.3245 | |
| 3-(N-Acetyl-L-cystein-S-yl)-acetaminophen | ||||
| Gestational age ≥37wks | N | 522 | 40 | 104 |
| OR | 1 (Ref.) | 1.28 | 0.78 | |
| 95% CI | - | (0.76,2.16) | (0.53,1.15) | |
| P-value | - | 0.3558 | 0.2151 | |
| Gestational age <37wks | N | 91 | 15 | 38 |
| OR | 1 (Ref.) | 1.72 | 1.29 | |
| 95% CI | - | (0.75,3.94) | (0.69,2.38) | |
| P-value | - | 0.2023 | 0.4243 | |
Abbreviations: N, number of participants; OR, odds ratio; 95% CI, 95% confidence interval. ORs were estimated from multinomial regression models in which each continuous acetaminophen biomarker (ranked inverse transformed) was an independent variable. Low birth weight was defined as birth weight <2500g. The adjusted model includes the following covariates: sex, gestational age at birth, primiparous parity, delivery type, maternal race and ethnicity, maternal age, maternal education, maternal marital status, prenatal smoking, maternal pre-pregnancy BMI, any gestational conditions, intrauterine inflammation, maternal asthma, maternal stress during pregnancy
eTable 6.
Distribution of 3-(N-Acetyl-L-cystein-S-yl)-acetaminophen value (inverse normal transformed) by count of acetaminophen biomarkers detected
| 3-(N-Acetyl-L-cystein-S-yl)-acetaminophen (inverse normal transformed) | ||||||
|---|---|---|---|---|---|---|
| Count of acetaminophen biomarkers detected | Number of participants | Minimum | 25th percentile | Median | 75th percentile | Maximum |
| 1 | 92 | −0.36 | 0.63 | 0.72 | 0.82 | 1.19 |
| 2 | 25 | 0.57 | 0.93 | 1.02 | 1.06 | 1.20 |
| 3 | 121 | 0.89 | 1.22 | 1.44 | 1.78 | 3.23 |
eFigure 1. Histograms of the raw intensities of acetaminophen metabolites from liquid chromatography-tandem mass spectrometry (LC-MS).
The horizontal axis shows the raw intensity of each acetaminophen metabolite on the log10 scale. The vertical axis shows the frequency for the specific bin. For each metabolite, 100 bins were created across the range of the raw intensity from LC-MS measurement. Red dotted line shows the background noise level for each metabolite. Values that were detected by LC-MS but below the noise level were colored gray. Values above the noise level were colored black.
eFigure 2. Pathways for acetaminophen metabolism.
Acetaminophen is mainly metabolized in liver through three pathways: glucuronidation, sulfation, and phase I oxidation. Glutathione (GSH) detoxifies NAPQI, and the process is catalyzed by GSH S-transferase. Depletion of GSH can lead to cell injury. A small proportion of acetaminophen remain. The chemical structures of the metabolites were obtained from The Human Metabolome Database (HMDB) (http://hmdb.ca).
Footnotes
Conflict of Interest: All other authors declare that they have no conflict of interest.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Dubovyi A, Chelimo C, Schierding W, Bisyuk Y, Camargo CA, Grant CC. A systematic review of asthma case definitions in 67 birth cohort studies. Paediatr Respir Rev. 2021;37:89–98. doi: 10.1016/J.PRRV.2019.12.005 [DOI] [PubMed] [Google Scholar]
- 2.Akar-Ghibril N, Casale T, Custovic A, Phipatanakul W. Allergic Endotypes and Phenotypes of Asthma. J Allergy Clin Immunol Pract. 2020;8(2):429–440. doi: 10.1016/J.JAIP.2019.11.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Romanet-Manent S, Charpin D, Magnan A, Lanteaume A, Vervloet D. Allergic vs nonallergic asthma: what makes the difference? Allergy. 2002;57(7):607–613. doi: 10.1034/J.1398-9995.2002.23504.X [DOI] [PubMed] [Google Scholar]
- 4.Landgraf-Rauf K, Anselm B, Schaub B. The puzzle of immune phenotypes of childhood asthma. Mol Cell Pediatr 2016 31. 2016;3(1):1–10. doi: 10.1186/S40348-016-0057-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Janson C, Kalm-Stephens P, Foucard T, Alving K, Nordvall SL. Risk factors associated with allergic and non-allergic asthma in adolescents. Clin Respir J. 2007;1(1):16–22. doi: 10.1111/J.1752-699X.2007.00001.X [DOI] [PubMed] [Google Scholar]
- 6.Zhu Z, Hasegawa K, Camargo CA, Liang L. Investigating asthma heterogeneity through shared and distinct genetics: Insights from genome-wide cross-trait analysis. J Allergy Clin Immunol. 2021;147(3):796–807. doi: 10.1016/j.jaci.2020.07.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bauer AZ, Swan SH, Kriebel D, Liew Z, Taylor HS, Bornehag CG, et al. Paracetamol use during pregnancy — a call for precautionary action. Nat Rev Endocrinol. 2021;17(12):757–766. doi: 10.1038/s41574-021-00553-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Levy G, Garrettson LK, Soda DM. Evidence of Placental Transfer of Acetaminophen. Pediatrics. 1975;55(6):895-895. doi: 10.1542/PEDS.55.6.895 [DOI] [PubMed] [Google Scholar]
- 9.Etminan M, Sadatsafavi M, Jafari S, Doyle-Waters M, Aminzadeh K, FitzGerald JM. Acetaminophen use and the risk of asthma in children and adults: A systematic review and metaanalysis. Chest. 2009;136(5):1316–1323. doi: 10.1378/chest.09-0865 [DOI] [PubMed] [Google Scholar]
- 10.Cheelo M, Lodge CJ, Dharmage SC, Simpson JA, Matheson M, Heinrich J, et al. Paracetamol exposure in pregnancy and early childhood and development of childhood asthma: A systematic review and meta-analysis. Arch Dis Child. 2015;100(1):81–89. doi: 10.1136/archdischild-2012-303043 [DOI] [PubMed] [Google Scholar]
- 11.Scialli AR, Ang R, Breitmeyer J, Royal MA. Childhood asthma and use during pregnancy of acetaminophen. A critical review. Reprod Toxicol. 2010;30(4):508–519. doi: 10.1016/j.reprotox.2010.09.005 [DOI] [PubMed] [Google Scholar]
- 12.Eyers S, Weatherall M, Jefferies S, Beasley R. Paracetamol in pregnancy and the risk of wheezing in offspring: a systematic review and meta-analysis. Clin Exp Allergy. 2011;41(4):482–489. doi: 10.1111/J.1365-2222.2010.03691.X [DOI] [PubMed] [Google Scholar]
- 13.Eneli I, Sadri K, Camargo C, Barr RG. Acetaminophen and the Risk of Asthma: The Epidemiologic and Pathophysiologic Evidence. Chest. 2005;127(2):604–612. doi: 10.1378/CHEST.127.2.604 [DOI] [PubMed] [Google Scholar]
- 14.Flint RB, Roofthooft DW, Van Rongen A, Van Lingen RA, Van Den Anker JN, Van Dijk M, et al. Exposure to acetaminophen and all its metabolites upon 10, 15, and 20 mg/kg intravenous acetaminophen in very-preterm infants. Pediatr Res. 2017;82(4):678–684. doi: 10.1038/pr.2017.129 [DOI] [PubMed] [Google Scholar]
- 15.Wang X, Zuckerman B, Pearson C, Kaufman G, Chen C, Wang G, et al. Maternal cigarette smoking, metabolic gene polymorphism, and infant birth weight. J Am Med Assoc. 2002;287(2):195–202. doi: 10.1001/jama.287.2.195 [DOI] [PubMed] [Google Scholar]
- 16.Wang G, Divall S, Radovick S, Paige D, Ning Y, Chen Z, et al. Preterm birth and random plasma insulin levels at birth and in early childhood. jamanetwork.com. 2014;311(6):587–596. doi: 10.1001/jama.2014.1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Roberts LD, Souza AL, Gerszten RE, Clish CB. Targeted metabolomics. Curr Protoc Mol Biol. 2012;1(SUPPL.98). doi: 10.1002/0471142727.MB3002S98 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Zhang M, Buckley JP, Liang L, Hong X, Wang G, Wang MC, et al. A metabolome-wide association study of in utero metal and trace element exposures with cord blood metabolome profile: Findings from the Boston Birth Cohort. Environ Int. 2022;158:106976. doi: 10.1016/J.ENVINT.2021.106976 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Brookhuis SAM, Allegaert K, Hanff LM, Lub-De Hooge MN, Dallmann A, Mian P. Modelling Tools to Characterize Acetaminophen Pharmacokinetics in the Pregnant Population. Pharm 2021 Vol 13 Page 1302. 2021;13(8):1302. doi: 10.3390/PHARMACEUTICS13081302 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Miners J, Robson R, Birkett D. Paracetamol metabolism in pregnancy. Br J Clin Pharmacol. 1986;22(3):359–362. doi: 10.1111/J.1365-2125.1986.TB02901.X [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Rayburn W, Shukla U, Stetson P, Piehl E. Acetaminophen pharmacokinetics: comparison between pregnant and nonpregnant women. Am J Obstet Gynecol. 1986;155(6):1353–1356. doi: 10.1016/0002-9378(86)90173–0 [DOI] [PubMed] [Google Scholar]
- 22.Nitsche JF, Patil AS, Langman LJ, Penn HJ, Derleth D, Watson WJ, et al. Transplacental Passage of Acetaminophen in Term Pregnancy. Am J Perinatol. 2017;34(6):541–543. doi: 10.1055/S-0036-1593845/ID/JR160291-16 [DOI] [PubMed] [Google Scholar]
- 23.Weatherall M, Ioannides S, Braithwaite I, Beasley R. The association between paracetamol use and asthma: Causation or coincidence? Clin Exp Allergy. 2015;45(1):108–113. doi: 10.1111/cea.12410 [DOI] [PubMed] [Google Scholar]
- 24.Shaheen SO, Lundholm C, Brew BK, Almqvist C. Prescribed analgesics in pregnancy and risk of childhood asthma. Eur Respir J. 2019;53(5):1–9. doi: 10.1183/13993003.01090-2018 [DOI] [PubMed] [Google Scholar]
- 25.Hodgman MJ, Garrard AR. A Review of Acetaminophen Poisoning. Crit Care Clin. 2012;28(4):499–516. doi: 10.1016/j.ccc.2012.07.006 [DOI] [PubMed] [Google Scholar]
- 26.Chen TS, Richie JP, Lang CA. Life span profiles of glutathione and acetaminophen detoxification. Drug Metab Dispos. 1990;18(6). [PubMed] [Google Scholar]
- 27.Micheli L, Cerretani D, Fiaschi AI, Giorgi G, Romeo MR, Runci FM. Effect of acetaminophen on glutathione levels in rat testis and lung. Environ Health Perspect. 1994;102(SUPPL. 9):63–64. doi: 10.1289/EHP.94102S963 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Dobrinskikh E, Sherlock LG, Orlicky DJ, Zheng L, de Dios R, Balasubramaniyan D, et al. The developing murine lung is susceptible to acetaminophen toxicity. Am J Physiol - Lung Cell Mol Physiol. 2021;320(5):L969–L978. doi: 10.1152/AJPLUNG.00072.2021/ASSET/IMAGES/LARGE/AJPLUNG.00072.2021_F004.JPEG [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Seidman DS, Laor A, Gale R, Stevenson DK, Danon YL. Is low birth weight a risk factor for asthma during adolescence? Arch Dis Child. 1991;66(5):584–587. doi: 10.1136/ADC.66.5.584 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Xu XF, Li YJ, Sheng YJ, Liu JL, Tang LF, Chen ZM. Effect of low birth weight on childhood asthma: A meta-analysis. BMC Pediatr. 2014;14(1):1–8. doi: 10.1186/1471-2431-14-275/FIGURES/5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Wjst M, Popescu M, Trepka MJ, Heinrich J, Wichmann HE. Pulmonary function in children with initial low birth weight. Pediatr Allergy Immunol. 1998;9(2):80–90. doi: 10.1111/J.1399-3038.1998.TB00308.X [DOI] [PubMed] [Google Scholar]
- 32.Boehm G, Müller DM, Teichmann B, Krumbiegel P. Influence of intrauterine growth retardation on parameters of liver function in low birth weight infants. Eur J Pediatr 1990 1496. 1990;149(6):396–398. doi: 10.1007/BF02009657 [DOI] [PubMed] [Google Scholar]
- 33.Soo JY, Wiese MD, Berry MJ, McMillen IC, Morrison JL. Intrauterine growth restriction may reduce hepatic drug metabolism in the early neonatal period. Pharmacol Res. 2018;134:68–78. doi: 10.1016/J.PHRS.2018.06.003 [DOI] [PubMed] [Google Scholar]
- 34.McBride GM, Meakin AS, Soo JY, Darby JRT, Varcoe TJ, Bradshaw EL, et al. Intrauterine growth restriction alters the activity of drug metabolising enzymes in the maternal-placental-fetal unit. Life Sci. 2021;285:120016. doi: 10.1016/J.LFS.2021.120016 [DOI] [PubMed] [Google Scholar]
- 35.Riece K, Yiong Huak C, Teng Nging T, Van Bever HP. A matched patient-sibling study on the usage of paracetamol and the subsequent development of allergy and asthma. Pediatr Allergy Immunol. 2007;18(2):128–134. doi: 10.1111/j.1399-3038.2006.00484.x [DOI] [PubMed] [Google Scholar]
- 36.Shaheen SO, Newson RB, Sherriff A, Henderson AJ, Heron JE, Burney PGJ, et al. Paracetamol use in pregnancy and wheezing in early childhood. Thorax. 2002;57(11):958–963. doi: 10.1136/thorax.57.11.958 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Shaheen SO, Newson RB, Henderson AJ, Headley JE, Stratton FD, Jones RW, et al. Prenatal paracetamol exposure and risk of asthma and elevated immunoglobulin E in childhood. Clin Exp Allergy. 2005;35(1):18–25. doi: 10.1111/j.1365-2222.2005.02151.x [DOI] [PubMed] [Google Scholar]
- 38.Servey J, physician JCA family, 2014 undefined. Over-the-counter medications in pregnancy. aafp.org. Accessed March 5, 2022. https://www.aafp.org/afp/2014/1015/p548.html?fbclid=IwAR3yURsFteKrdM_WTG0yqLtXugQ1cbOrGe4S9LW7Wv_eoSL8BykX9eTYSHE [PubMed]
- 39.FDA recommends avoiding use of NSAIDs in pregnancy at 20 weeks or later because they can result in low amniotic fluid | FDA. Accessed March 2, 2022. https://www.fda.gov/drugs/drug-safety-and-availability/fda-recommends-avoiding-use-nsaids-pregnancy-20-weeks-or-later-because-they-can-result-low-amniotic [Google Scholar]
- 40.O’Sullivan J, Cairns AE, Plesca E, Black RS, Frise C, Vatish M, et al. Paracetamol use in pregnancy — neglecting context promotes misinterpretation. Nat Rev Endocrinol 2022 186. 2022;18(6):385-385. doi: 10.1038/s41574-022-00656-9 [DOI] [PubMed] [Google Scholar]
References
- 1.Roberts LD, Souza AL, Gerszten RE, Clish CB. Targeted metabolomics. Curr Protoc Mol Biol. 2012;1(SUPPL.98). doi: 10.1002/0471142727.MB3002S98 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Zhang M, Buckley JP, Liang L, Hong X, Wang G, Wang MC, et al. A metabolome-wide association study of in utero metal and trace element exposures with cord blood metabolome profile: Findings from the Boston Birth Cohort. Environ Int. 2022;158:106976. doi: 10.1016/J.ENVINT.2021.106976 [DOI] [PMC free article] [PubMed] [Google Scholar]