Introduction: Preterm neonates often receive a variety of duration of antibiotic exposure during admission. The aim of the study was to evaluate whether neonatal antibiotic exposure is relevant with longitudinal growth problems in preterm-birth children. Methods: This prospective study enrolled 481 infants who were born <32 weeks of gestation, discharged, and longitudinally followed from corrected age (CA) 6–60 months. After excluding 153 infants with blood culture-confirmed bacteremia, necrotizing enterocolitis, severe cerebral palsy, intestinal ostomy, and congenital anomaly, 328 infants were included for analysis. Covariates included perinatal demographics, neonatal morbidities, extrauterine growth restriction, and antibiotic exposure accumulated by term equivalent age. The primary outcome was the anthropometric trajectories in z-score of bodyweight (zBW), body height (zBH), and body mass index (zBMI) from CA 6–60 months. Results: Antibiotic exposure duration was significantly negatively associated with zBW and zBH at CA 6, 12, and 60 months, and zBMI at CA 60 months. Multivariate generalized estimating equation analyses showed antibiotic exposure duration had significantly faltering z-score increment from CA 6 to 60 months in zBW and zBH (adjusted mean [95% CI]; ΔzBW: −0.021 [−0.041 to −0.001], p = 0.042; ΔzBH: −0.019 [−0.035 to −0.002], p = 0.027) after adjustment. Children with neonatal antibiotic exposure duration >15 days were significantly lower in the mean anthropometric zBW, zBH, and zBMI at CA 6, 12, 24, and 60 months compared with children with neonatal antibiotic exposure ≤15 days (all p < 0.01). Conclusions: Growth increments were negatively associated with antibiotic exposure duration in preterm neonates implicating that antibiotic stewardship and growth follow-up for preterm neonates are thus warranted.

Preterm infants often receive a variety of antibiotic exposure duration during admission. This study was to evaluate whether neonatal antibiotic exposure is associated with longitudinal growth problems in preterm-birth children. This study followed a cohort of 328 very preterm infants born <32 weeks of gestation, who were discharged from a tertiary hospital without documented neonatal bacterial infection from 2004 to 2016 and received longitudinal follow-up for growth outcomes at their corrected age (CA) 6, 12, 24, and 60 months. We collected perinatal demographics, neonatal morbidities, extrauterine growth restriction, and antibiotic exposure duration by term equivalent age, to associate with trajectories in z-scores of bodyweight (zBW), body height (zBH), and body mass index (zBMI) from CA 6–60 months. We found antibiotic exposure duration was negatively associated with zBW and zBH at CA 6, 12, and 60 months, and zBMI at CA 60 months. After adjustments, antibiotic exposure duration was associated with decreased growth velocity in zBW and zBH from CA 6–60 months. Children with longer antibiotic exposure >15 days had significantly lower zBW, zBH, and zBMI at CA 6, 12, 24, and 60 months compared with children with antibiotic exposure ≤15 days (all p < 0.01), suggesting that longer antibiotic exposure duration was associated with inferior weight and height gains. This study concludes that long-term growth increments may negatively associate with neonatal antibiotic exposure duration in children born very premature, implicating that antibiotic stewardship and growth follow-up for preterm neonates are thus warranted.

In recent decades, the prevalence of overweight and obese children has risen substantially worldwide [1, 2]. Obesity in children is often associated with development of metabolic syndrome and cardiovascular risks in later life. Antibiotic exposure in early life that changes the gut microbiome development may potentially impact the long-term childhood health [3, 4]. Studies including meta-analyses have shown evidence that antibiotic exposure in early life significantly increases the risk of childhood weight gain and obesity [5‒7].

Studies on antibiotic exposure and growth outcome changes in children have been mainly focused on term-birth neonates. In contrast to the uncommon use of antibiotics in term-birth neonates, very preterm-birth neonates of gestational age less than 32 weeks often have a variable duration of antibiotic exposure in the neonatal intensive care unit (NICU) even without documented evidence of bacterial infections [8, 9]. Obesity is also an important epidemic problem in preterm-birth children [10, 11]. Whether antibiotic exposure in the neonatal period associates with the overweight or growth problem in very preterm-birth children remains unknown.

It is also important to differentiate between the effects of antibiotic use in infancy and that of underlying infections on the risk of childhood growth alterations [12]. It remains unclear whether growth outcomes are affected by antibiotic exposure duration in preterm neonates without documented evidence of bacterial infections. Using the prospective registry data of longitudinal growth outcomes at corrected ages (CAs) 6, 12, 24, and 60 months of very preterm neonates without documented bacterial infections, this study aimed to examine the associations of antibiotic exposure duration in the neonatal period with the longitudinal anthropometric trajectories in z-score of bodyweight (zBW), body height (zBH), and body mass index (zBMI) from CA 6 to 60 months in preterm-birth children.

Study Population

A total of 696 very preterm-birth infants who were born with gestational age less than 32 weeks were admitted within 24 h after birth to the university hospital between January 2004 and September 2016 (shown in Fig. 1). Among the 595 infants who survived to discharge, 4 infants died after discharge. Of the 591 surviving infants after discharge, 110 infants lost to follow-up, and 481 (81.4%) infants received follow-up examinations by 60 months of CA. After reviewing the clinical history of the cohort, we excluded 16 children with severe cerebral palsy or prolonged intestinal ostomy, 3 with genetic/congenital conditions (1 with trisomy 18, 1 with Prader-Willi syndrome, and 1 with congenital brain malformation), and 134 with documented infection (culture-confirmed infection bacteremia or meningitis or necrotizing enterocolitis [NEC] ≥ stage II). We included 328 children for analysis. This cohort study was approved by the university hospital’s Institutional Review Board (No. ER-98-135), and informed consent was obtained from the parents of each infant during hospitalization and at follow-up visits.

Fig. 1.

Enrollment, exclusion, and inclusion for analysis of a longitudinal follow-up growth cohort of very preterm-birth children.

Fig. 1.

Enrollment, exclusion, and inclusion for analysis of a longitudinal follow-up growth cohort of very preterm-birth children.

Close modal

Antibiotic Exposure

Antibiotic exposure in the neonatal period was calculated based on the cumulative duration of antibiotics treatment from birth to postmenstrual age 42 weeks. The antibiotic exposure information was retrieved from the electronic medical system, and the calendar days during which antibiotics were prescribed were collected. Two or three antibiotics administered simultaneously on the same day were calculated as one exposure day. The class of antibiotics administered included aminopenicillins, cephalosporin, glycopeptide, aminoglycoside, carbapenem, metronidazole, etc. (online suppl. Table S1; for all online suppl. material, see https://doi.org/10.1159/000535946). Each course of antibiotic treatment and the related infection events were reviewed by two neonatologists. The antibiotic guideline used for very preterm neonates in the NICU included empiric therapy of ampicillin plus gentamicin for 2–3 days after septic workup on admission, and ampicillin/sulbactam plus amikacin for any suspected nosocomial infection, which were then stewarded according to the antibiotic sensitivity by bacteria culture.

Outcomes

The primary outcomes included extrauterine growth restriction (EUGR) at the term equivalent age (TEA), and the anthropometric trajectories of zBW, zBH, and zBMI measured at CA 6, 12, 24, and 60 months. A delta zBW from birth to TEA of −1 or less indicated EUGR [13]. The anthropometric zBW, zBH, and zBMI after TEA were calculated by the standard established by the World Health Organization [14]. The longitudinally measured anthropometric increments were used to model the growth trajectories for the changes in zBW, zBH, and zBMI from CA 6 months to 60 months.

Covariates

Demographics in the perinatal period, such as socioeconomic status, multiple gestations, clinical chorioamnionitis, BW at birth and sex, and neonatal major morbidities, including respiratory distress syndrome requiring surfactant therapy, patent ductus arteriosus requiring surgery, bronchopulmonary dysplasia (BPD) [15], postnatal steroid use for BPD, sepsis, NEC, severe intraventricular hemorrhage (IVH), cystic periventricular leukomalacia, and severe retinopathy of prematurity requiring surgery were identified. Small for gestational age (SGA) was defined as birth BW of less than 10th percentile for the gestational age [13]. The sociodemographic status determined by parents’ education and occupation was categorized into five classes (I–V), with class IV and class V as low socioeconomic status [16].

Statistical Analysis

All analyses were performed using SPSS (v.26, IBM, Armonk, NY, USA). The dependence of anthropometric variables on the clinical risks that might result in growth restriction, chosen a priori, and antibiotic exposure was first assessed using univariate analysis. Linear regression analysis and logistic regression were used for scaled dependent variables and binary dependent variables in anthropometric measurements, respectively, to obtain coefficient (β) and odds ratios of independent variables. Covariates were selected for the multivariable analysis after taking into account the significance obtained from the univariate analysis by p < 0.1, clinical relevance, and collinearity between the chosen variables. Moreover, a generalized estimating equation (GEE) was used to analyze the association between repeated measurements and factors. In the GEE model, a clinical risk having significant coefficients with an anthropometric variable that repeatedly measured at the four follow-up visits was interpreted as a factor associated with growth increments. Graphic trajectories were illustrated using GraphPad Prism (version 9.0.0 for Windows, GraphPad Software, San Diego, CA, USA). The results were considered statistically significant if a p value <0.05.

Effects of Neonatal Antibiotic Exposure Duration on EUGR by TEA

Table 1 shows the demographics, neonatal morbidities, status at TEA, and the follow-up anthropometric z-scores of the 328 very preterm neonates (gestational age, mean ± SD, 28.2 ± 2.0 weeks) without documented bacterial infection. The mean duration of antibiotic exposure was 10.7 ± 9.4 days with an interquartile range of 4–15 days. By TEA, a total of 121 (37%) neonates had EUGR at TEA. The univariate model followed by multivariate logistic regression analysis showed that EUGR at TEA was significantly associated with the antibiotic exposure duration (aOR: 1.108, 95% confidence interval: 1.062–1.156, p < 0.001) (online suppl. Table S2).

Table 1.

The demographics, neonatal morbidities, and growth status at TEA of 328 very preterm-birth neonates without documented bacterial infection

Total infants, N = 328
At birth 
 Male, n (%) 151 (46) 
 Gestational age, mean±SD, weeks 28.2±2.0 
 Bodyweight, mean±SD, grams 1076±252 
 Bodyweight, z-scores, mean±SD −0.33±0.79 
 SGA, n (%) 42 (13) 
 Multi-gestation, n (%) 104 (32) 
 Low socioeconomic status, n (%) 95 (29) 
Neonatal morbidities 
 RDS requiring surfactant, n (%) 72 (22) 
 PDA requiring surgery, n (%) 39 (12) 
 Severe IVH, n (%) 8 (2) 
 cPVL, n (%) 13 (4) 
 BPD, n (%) 89 (27) 
 Postnatal steroid use, n (%) 24 (7) 
 ROP requiring surgery, n (%) 22 (7) 
Antibiotics utilization* 
 Empirical therapy at admission, n (%) 328 (100) 
 Antibiotic treatment, courses, mean±SD 2.9±2.2 
 Antibiotic exposure duration, days  
  Mean±SD 10.7±9.4 
  Median (interquartile range, Q1–Q3) 7 (4–15) 
Anthropometry at TEA 
 PMA of assessment, mean±SD, weeks 38.7±1.5 
 Bodyweight, mean±SD, grams 2,969±821 
 Bodyweight, z-scores, mean±SD −0.92±1.4 
 EUGR of bodyweight, n (%) 121 (37) 
Total infants, N = 328
At birth 
 Male, n (%) 151 (46) 
 Gestational age, mean±SD, weeks 28.2±2.0 
 Bodyweight, mean±SD, grams 1076±252 
 Bodyweight, z-scores, mean±SD −0.33±0.79 
 SGA, n (%) 42 (13) 
 Multi-gestation, n (%) 104 (32) 
 Low socioeconomic status, n (%) 95 (29) 
Neonatal morbidities 
 RDS requiring surfactant, n (%) 72 (22) 
 PDA requiring surgery, n (%) 39 (12) 
 Severe IVH, n (%) 8 (2) 
 cPVL, n (%) 13 (4) 
 BPD, n (%) 89 (27) 
 Postnatal steroid use, n (%) 24 (7) 
 ROP requiring surgery, n (%) 22 (7) 
Antibiotics utilization* 
 Empirical therapy at admission, n (%) 328 (100) 
 Antibiotic treatment, courses, mean±SD 2.9±2.2 
 Antibiotic exposure duration, days  
  Mean±SD 10.7±9.4 
  Median (interquartile range, Q1–Q3) 7 (4–15) 
Anthropometry at TEA 
 PMA of assessment, mean±SD, weeks 38.7±1.5 
 Bodyweight, mean±SD, grams 2,969±821 
 Bodyweight, z-scores, mean±SD −0.92±1.4 
 EUGR of bodyweight, n (%) 121 (37) 

SGA, small for the gestational age; RDS, respiratory distress syndrome; PDA, patent ductus arteriosus; IVH, intraventricular hemorrhage; cPVL, cystic periventricular leukomalacia; BPD, bronchopulmonary dysplasia; ROP, retinopathy of prematurity; PMA, postmenstrual age; EUGR, extrauterine growth restriction defined as a decrease in the zBW by 1 or more from birth to TEA and defined as a decrease in the zBW by −1 or more from birth to TEA; TEA: PMA 37–42 weeks; *antibiotic utilization was monitored by PMA 42 weeks.

Negative Associations of Neonatal Antibiotic Exposure Duration with the Anthropometric zBW, zBH, and zBMI at Follow-Up

The antibiotic exposure duration per day showed significantly negative associations with the zBW and zBH at CA 6, 12, and 60 months, and zBMI at CA 60 months (adjusted β at CA 60 months: zBW, −0.025; zBH, −0.016; zBMI, −0.023, all p < 0.05; a value of β equivalent to a z-score) after adjusting for gestational age, SGA, sex, and EUGR at TEA (shown in Table 2). The univariate GEE methods for trajectory analysis showed that SGA, EUGR at TEA, and neonatal antibiotic exposure duration were negatively associated with the increment trajectories of zBW, zBH, and zBMI from CA 6 months to 60 months (online suppl. Table S3). Severe IVH and postnatal steroid use for BPD were the additional risks associated with negative increment trajectory of zBW and zBMI. Further multivariate GEE analysis showed lower gestational age and EUGR at TEA were the shared risk factors associated with increments of zBW, zBH, and zBMI (shown in Table 3). Antibiotic exposure duration per day had significantly faltering increment effects in the trajectory of zBW and zBH (adjusted mean ΔzBW, −0.021, p = 0.042; adjusted mean ΔzBH, −0.019, p = 0.027; translated to Δgrams and Δcentimeters in online suppl. Fig. S1).

Table 2.

Negative associations between the duration of neonatal antibiotic exposure and the mean anthropometric zBW, zBH, and zBMI at CA 6, 12, 24, and 60 months

Raw scores, mean±SDCoefficients with the duration of antibiotic exposure
unadjustedadjusted*
β95% CIp valueβ95% CIp value
zBW 
 6 months −0.36±1.20 −0.031 −0.045, −0.016 <0.001 −0.024 −0.042, −0.006 0.010 
 12 months −0.50±1.11 −0.025 −0.038, −0.013 <0.001 −0.022 −0.038, −0.006 0.008 
 24 months −0.45±1.09 −0.020 −0.032, −0.007 0.002 −0.014 −0.030, 0.002 0.092 
 60 months −0.41±1.21 −0.023 −0.037, −0.009 0.001 −0.025 −0.043, −0.007 0.006 
zBH 
 6 months −0.37±1.13 −0.028 −0.043, −0.013 <0.001 −0.019 −0.038, −0.001 0.038 
 12 months −0.27±1.19 −0.023 −0.037, −0.009 0.001 −0.018 −0.035, 0.000 0.044 
 24 months −0.38±1.04 −0.015 −0.027, −0.003 0.012 −0.012 −0.028, 0.003 0.127 
 60 months −0.34±1.03 −0.016 −0.028, −0.005 0.006 −0.016 −0.031, −0.001 0.041 
zBMI 
 6 months −0.14±1.20 −0.020 −0.034, −0.006 0.006 −0.018 −0.036, 0.001 0.062 
 12 months −0.48±1.06 −0.017 −0.029, −0.005 0.002 −0.016 −0.032, 0.000 0.050 
 24 months −0.34±1.13 −0.015 −0.028, −0.002 0.001 −0.010 −0.027, 0.008 0.284 
 60 months −0.33±1.23 −0.019 −0.033, −0.005 0.001 −0.023 −0.041, −0.005 0.013 
Raw scores, mean±SDCoefficients with the duration of antibiotic exposure
unadjustedadjusted*
β95% CIp valueβ95% CIp value
zBW 
 6 months −0.36±1.20 −0.031 −0.045, −0.016 <0.001 −0.024 −0.042, −0.006 0.010 
 12 months −0.50±1.11 −0.025 −0.038, −0.013 <0.001 −0.022 −0.038, −0.006 0.008 
 24 months −0.45±1.09 −0.020 −0.032, −0.007 0.002 −0.014 −0.030, 0.002 0.092 
 60 months −0.41±1.21 −0.023 −0.037, −0.009 0.001 −0.025 −0.043, −0.007 0.006 
zBH 
 6 months −0.37±1.13 −0.028 −0.043, −0.013 <0.001 −0.019 −0.038, −0.001 0.038 
 12 months −0.27±1.19 −0.023 −0.037, −0.009 0.001 −0.018 −0.035, 0.000 0.044 
 24 months −0.38±1.04 −0.015 −0.027, −0.003 0.012 −0.012 −0.028, 0.003 0.127 
 60 months −0.34±1.03 −0.016 −0.028, −0.005 0.006 −0.016 −0.031, −0.001 0.041 
zBMI 
 6 months −0.14±1.20 −0.020 −0.034, −0.006 0.006 −0.018 −0.036, 0.001 0.062 
 12 months −0.48±1.06 −0.017 −0.029, −0.005 0.002 −0.016 −0.032, 0.000 0.050 
 24 months −0.34±1.13 −0.015 −0.028, −0.002 0.001 −0.010 −0.027, 0.008 0.284 
 60 months −0.33±1.23 −0.019 −0.033, −0.005 0.001 −0.023 −0.041, −0.005 0.013 

CI, confidence interval.

*Adjusted with gestational age, SGA, sex, and EUGR at TEA; β, linear regression coefficient; statistical significance of linear regressions were assumed for p < 0.05.

Table 3.

The risks associated with increments of the trajectories in zBW, zBH, and zBMI from CA 6 months–60 months: a multivariate GEE analysis

zBWzBHzBMI
Mean Δz95% CI lower, upperp valueMean Δz95% CI lower, upperp valueMean Δz95% CI lower, upperp value
No. of scheduled follow-up −0.011 −0.053, 0.030 0.590 −0.062 −0.098, −0.025 0.001 0.023 −0.023, 0.068 0.328 
Gestational age, weeks −0.107 −0.177, −0.037 0.003 −0.075 −0.138, −0.013 0.018 −0.099 −0.168, −0.030 0.005 
SGA −0.601 −0.875, −0.327 <0.001 −0.761 −1.078, −0.443 <0.001 −0.175 −0.431, 0.081 0.181 
Male 0.016 −0.188, 0.220 0.877 −0.125 −0.323, 0.074 0.218 0.138 −0.057, 0.333 0.166 
Low socioeconomic status NA NA NA 0.195 −0.026, 0.416 0.084 NA NA NA 
PDA requiring surgery NA NA NA 0.071 −0.267, 0.409 0.680 NA NA NA 
Severe IVH −0.589 −1.205, 0.028 0.061 NA NA NA −1.028 −1.681, −0.374 0.002 
BPD NA NA NA −0.036 −0.339, 0.267 0.817 −0.079 −0.364, 0.206 0.587 
Postnatal steroid for BPD −0.100 −0.538, 0.339 0.656 NA NA NA −0.231 −0.642, 0.217 0.331 
EUGR at TEA (ΔzBW <–1) −0.613 −0.874, −0.352 <0.001 −0.426 −0.669, −0.184 0.001 −0.486 −0.737, −0.235 <0.001 
Antibiotic exposure duration, days −0.021 −0.041, −0.001 0.042 −0.019 −0.035, −0.002 0.027 −0.012 −0.031, 0.008 0.240 
zBWzBHzBMI
Mean Δz95% CI lower, upperp valueMean Δz95% CI lower, upperp valueMean Δz95% CI lower, upperp value
No. of scheduled follow-up −0.011 −0.053, 0.030 0.590 −0.062 −0.098, −0.025 0.001 0.023 −0.023, 0.068 0.328 
Gestational age, weeks −0.107 −0.177, −0.037 0.003 −0.075 −0.138, −0.013 0.018 −0.099 −0.168, −0.030 0.005 
SGA −0.601 −0.875, −0.327 <0.001 −0.761 −1.078, −0.443 <0.001 −0.175 −0.431, 0.081 0.181 
Male 0.016 −0.188, 0.220 0.877 −0.125 −0.323, 0.074 0.218 0.138 −0.057, 0.333 0.166 
Low socioeconomic status NA NA NA 0.195 −0.026, 0.416 0.084 NA NA NA 
PDA requiring surgery NA NA NA 0.071 −0.267, 0.409 0.680 NA NA NA 
Severe IVH −0.589 −1.205, 0.028 0.061 NA NA NA −1.028 −1.681, −0.374 0.002 
BPD NA NA NA −0.036 −0.339, 0.267 0.817 −0.079 −0.364, 0.206 0.587 
Postnatal steroid for BPD −0.100 −0.538, 0.339 0.656 NA NA NA −0.231 −0.642, 0.217 0.331 
EUGR at TEA (ΔzBW <–1) −0.613 −0.874, −0.352 <0.001 −0.426 −0.669, −0.184 0.001 −0.486 −0.737, −0.235 <0.001 
Antibiotic exposure duration, days −0.021 −0.041, −0.001 0.042 −0.019 −0.035, −0.002 0.027 −0.012 −0.031, 0.008 0.240 

CI, confidence interval; SGA, small for gestational age; severe IVH, grade III and IV IVH; EUGR, extrauterine growth restriction and defined as a decrease in the zBW by 1 or more from birth to TEA; TEA, term equivalent age (PMA 37–42 weeks); GEE, generalized estimating equation; Δz, increment in the z-score from 6 months to 60 months; PMA, postmenstrual age; PDA, patent ductus arteriosus.

Prolonged Neonatal Antibiotic Exposure Associated with Lower Mean Anthropometric zBW, zBH, and zBMI at CA 6, 12, 24, and 60 Months

Based on the quartile of neonatal antibiotic exposure duration, we further categorized the children into two groups: children whose neonatal antibiotic exposure was in the first to third quartile (Q1–Q3; antibiotic exposure ≤15 days) and children whose neonatal antibiotic exposure was in the fourth quartile (Q4; antibiotic exposure >15 days). After adjustment with gestational age, sex, SGA, severe IVH, postnatal steroid use for BPD, and EUGR at TEA, the children with neonatal antibiotic exposure >15 days were significantly lower in the mean anthropometric zBW, zBH, and zBMI at CA 6, 12, 24, and 60 months compared with the children with neonatal antibiotic exposure ≤15 days (all p < 0.01) (shown in Fig. 2).

Fig. 2.

Differences of anthropometric mean zBW, zBH, and zBMI at CA 6 months, 12 months, 24 months, and 60 months between children whose neonatal antibiotic exposure duration was in the first to third quartile (Q1–Q3; antibiotic exposure ≤15 days) and children whose neonatal antibiotic exposure duration was in the fourth quartile (Q4; antibiotic exposure >15 days). Upper panel (a, b, c) for unadjusted z-score; lower panel (d, e, f) for adjusted z-score; adjustment with gestational age, sex, SGA, severe IVH, postnatal steroid use for BPD, and extrauterine growth retardation at TEA. Blue line: for children with neonatal antibiotic exposure duration ≤15 days; red line for children with neonatal antibiotic exposure duration >15 days. Data are presented as mean ± SEM. Statistical significance of p values: ** <0.01, *** <0.001.

Fig. 2.

Differences of anthropometric mean zBW, zBH, and zBMI at CA 6 months, 12 months, 24 months, and 60 months between children whose neonatal antibiotic exposure duration was in the first to third quartile (Q1–Q3; antibiotic exposure ≤15 days) and children whose neonatal antibiotic exposure duration was in the fourth quartile (Q4; antibiotic exposure >15 days). Upper panel (a, b, c) for unadjusted z-score; lower panel (d, e, f) for adjusted z-score; adjustment with gestational age, sex, SGA, severe IVH, postnatal steroid use for BPD, and extrauterine growth retardation at TEA. Blue line: for children with neonatal antibiotic exposure duration ≤15 days; red line for children with neonatal antibiotic exposure duration >15 days. Data are presented as mean ± SEM. Statistical significance of p values: ** <0.01, *** <0.001.

Close modal

Antibiotic treatment in very preterm infants is often empirically prescribed, and unnecessary and prolonged antibiotic exposure has been documented in numerous studies. This study used a longitudinal growth follow-up cohort of very preterm-birth children without documented neonatal bacterial infections to address the duration effects of neonatal antibiotic exposure on the anthropometric growth of zBW, zBH, and zBMI from CA 6 to 60 months. Our study showed that the duration of neonatal antibiotic exposure significantly contributed to EUGR at TEA. Importantly, antibiotic exposure duration per day in neonatal stage showed significantly negative associations with the lower mean zBW and zBH at CA 6, 12, and 60 months. Each additional day of antibiotic exposure was associated with slower increments in anthropometric zBW and zBH from CA 6 to 60 months. From the analysis by the quartile duration of neonatal antibiotic exposure, children with antibiotic exposure at Q4 (antibiotic exposure >15 days) had significantly lower mean zBW, zBH, and zBMI at CA 6, 12, 24 and 60 months than children with antibiotic exposure at Q1–Q3 (antibiotic exposure ≤15 days). These findings suggest that the duration of neonatal antibiotic exposure is negatively associated not only with EUGR by TEA but also with long-term longitudinal growth increment trajectories of very preterm-birth children without neonatal bacterial infections (online suppl. Fig. S1).

Most studies found a positive relationship between early-life antibiotic exposure and overweight or obesity outcomes at follow-up in children. However, very few studies have differentiated between the effects of antibiotic exposure and underlying infectious diseases on the risk of childhood obesity [5, 6, 12, 17]. One birth cohort study in mostly term-birth infants found that after controlling for confounding factors including preterm delivery, infection without antibiotic use in infancy was associated with an increased risk of childhood obesity compared with controls without infection. By contrast, compared with infants with untreated infections, antibiotic exposure during infancy was not associated with childhood obesity, suggesting that infections but not antibiotic exposure during infancy are associated with a risk of childhood obesity [12]. To distinguish between the impact of antibiotic exposure and underlying infectious diseases, this study examined the duration effects of neonatal antibiotic exposure on longitudinal growth outcomes in very preterm children without confirmed neonatal bacterial infection.

In the neonatal period, most term-birth neonates are not exposed to antibiotics unless they are suspected of having neonatal sepsis. In contrast, almost every very preterm-birth neonate is exposed to a variable duration of antibiotics in the NICU [8, 18]. While most studies investigate the effects of early-life antibiotic exposure on growth outcomes in term-birth infants [3, 5, 19], few have examined these effects in preterm-birth children [20‒22]. In infants born at 30–32 weeks of gestation, Reid et al. [22] characterized the relationship between the duration of antibiotic exposure in the first week of life and subsequent growth velocity during hospitalization. They found no distinction in the zBW among the groups with no antibiotics, <5 days of antibiotics, or ≥5 days of antibiotics. Using a regression in preterm infants ≤32 weeks of gestation, Pyle et al. [21] showed that infants who were exposed to a median of 11 antibiotic days of therapy were not associated with weight or length delta z-scores from birth through CA 12 months. Here, we showed that neonatal antibiotic exposure duration is associated not only with reduced extrauterine growth by TEA but also with decreased growth increment trajectories in the first 5 years of life of very preterm children. These findings highlight that an antibiotic stewardship program in preterm neonates in the NICU is necessary not only to prevent the emergence of resistant bacteria species but also to protect against the adverse effects of antibiotic exposure on longitudinal growth outcomes [8, 23].

Neonatal morbidities, such as SGA, NEC, BPD, and EUGR, have been closely linked to long-term growth outcomes of children born very prematurely [24, 25]. However, few studies have considered these neonatal morbidity factors when examining the effects of neonatal or infantile antibiotic exposure on growth outcomes [3, 21]. We used GEE multivariable analysis to show that in addition to gestational age, other neonatal variables such as EUGR and antibiotic exposure duration were negatively associated with the growth trajectory increments of the z-scores in the first 5 years of life. No collinearity between SGA and antibiotic exposure was found (Spearman’s rank correlation = 0.016, p = 0.771). Together, these findings suggest the importance of including antibiotic exposure duration in the growth outcome study of preterm-birth children.

Early-life antibiotic exposure studies that focused on the infantile period or the first 12, 24, 48 months or 6 years of life in mostly term-birth children have shown increased bodyweight or obesity at follow-up [5, 17, 26]. In contrast, studies of neonatal antibiotic exposure in term-birth neonates revealed decreased weight gain or obesity in childhood [3, 27]. Our study showed negative association with long-term growth trajectory increments after neonatal antibiotic exposure. These findings that different critical time points of antibiotic exposure in early life have opposite growth outcomes suggest different metabolic programming effects may be induced by alterations in gut microbiota in response to the timing and duration of early-life antibiotic exposure in children [3, 28].

None of our study subjects had received caffeine treatment during admission. We used aminophylline and theophylline as the alternative of caffeine for apnea. We did not find association between exposure of aminophylline/theophylline and long-term growth outcomes. In addition, a large randomized trial had showed that the mean percentiles for height, weight, and head circumference at follow-up did not differ significantly between the caffeine and placebo groups [29].

The disease severity assessed by the Clinical Risk Index for Babies II (CRIB II) score at admission was associated with antibiotic exposure duration (Pearson correlation: 0.604; p < 0.001). However, the collinearity between antibiotics exposure duration and the disease severity was explained by gestation ages. After adjusting with CRIB II score and other parameters, we still find significant negative associations between antibiotic exposure duration and the mean anthropometric zBW, zBH, and zBMI at most of the follow-up visits (online suppl. Table S4). Meanwhile, after adjusting with CRIB II score in the multivariate GEE analysis, we also find significant negative associations between antibiotic exposure duration and anthropometric z-score increments of zBW, zBH, and zBMI (online suppl. Table S5).

This study is the first to examine the longitudinal growth trajectories in very preterm-birth children after neonatal antibiotic exposure. We did not specifically analyze the specific microbial alterations that associated with the growth trajectory changes after neonatal antibiotic exposure. The differences in the growth outcome after narrow- and broad-spectrum neonatal antibiotic exposure also remain to be determined. The rate of infection in very preterm infants in the NICU might be underestimated in those whose infectious conditions were not documented by positive blood cultures. Accurate identification of neonatal sepsis is challenging. Blood culture has been considered as a gold standard method but sepsis identification is complicated by a high false-negative results. Acute-phase reactants, cytokines, and cell surface antigens have been investigated as indicators for neonatal sepsis, but none of them are currently in routine clinical setting. A multicenter prospective study on the long-term growth outcome of antibiotic exposure in very preterm neonates may validate our findings.

Longitudinal growth trajectory in the first 5 years after birth was negatively associated with antibiotic exposure duration in the neonatal period of very preterm-birth children without documented bacterial infections. Antibiotic stewardship in the NICU and growth follow-up for preterm neonates are thus warranted.

The corresponding author Dr. Chao-Ching Huang had full access to the dataset used and analyzed during the current study. This manuscript was edited by the Foreign Language Center, National Cheng Kung University, Tainan, Taiwan. The authors thank the Taiwan Premature Baby Foundation and all team members in charge of data collection and assessment of the children. None of these individuals were compensated for their contributions.

The study was approved by the Institutional Review Board of National Cheng Kung University Hospital (ER-98-135) for the collection of parental information and neonatal data. The parents signed the informed consents when infants were hospitalized in the NICUs. The study was conducted in compliance with the Helsinki Declaration.

The authors have no conflicts of interest to declare.

This study was supported by Grants from the Taiwan National Science and Technology Counsel (NSTC-111-2314-B-006-001; 111-2314-B-006–012) and National Cheng Kung University Hospital (NCKUH-11201003). The funders had no role in designing or conducting the study; collecting, managing, analyzing, or interpreting the data; preparing, reviewing, or approving the manuscript; or deciding to submit the manuscript for publication.

Dr. Yung-Chieh Lin designed and conducted the study and drafted the manuscript. Dr. Chi-Hsiang Chu and Dr. Yen-Kuang Lin performed the statistical analyses, interpreted the data, and revised the manuscript for important intellectual content. Dr. Chih-Chia Chen and Li-Wen Chen assisted in data collection and illustration, and revised the manuscript for important intellectual content. Dr. Chao-Ching Huang conceptualized and designed the study, curated the data, and revised the manuscript. All authors approved the final manuscript as submitted and have agreed to be accountable for all aspects of the work.

All data used and analyzed during this study are included in this article and its online supplementary material. Further inquiries can be directed to the corresponding author.

1.
Dutton
H
,
Doyle
MA
,
Buchan
CA
,
Mohammad
S
,
Adamo
KB
,
Shorr
R
, et al
.
Antibiotic exposure and risk of weight gain and obesity: protocol for a systematic review
.
Syst Rev
.
2017
;
6
(
1
):
169
. .
2.
Lobstein
T
,
Jackson-Leach
R
,
Moodie
ML
,
Hall
KD
,
Gortmaker
SL
,
Swinburn
BA
, et al
.
Child and adolescent obesity: part of a bigger picture
.
Lancet
.
2015
;
385
(
9986
):
2510
20
. .
3.
Uzan-Yulzari
A
,
Turta
O
,
Belogolovski
A
,
Ziv
O
,
Kunz
C
,
Perschbacher
S
, et al
.
Neonatal antibiotic exposure impairs child growth during the first six years of life by perturbing intestinal microbial colonization
.
Nat Commun
.
2021
;
12
(
1
):
443
. .
4.
Yassour
M
,
Vatanen
T
,
Siljander
H
,
Hämäläinen
AM
,
Härkönen
T
,
Ryhänen
SJ
, et al
.
Natural history of the infant gut microbiome and impact of antibiotic treatment on bacterial strain diversity and stability
.
Sci Transl Med
.
2016
;
8
(
343
):
343ra81
. .
5.
Chelimo
C
,
Camargo
CA
Jr.
,
Morton
SMB
,
Grant
CC
.
Association of repeated antibiotic exposure up to age 4 Years with body mass at age 4.5 years
.
JAMA Netw Open
.
2020
;
3
(
1
):
e1917577
. .
6.
Baron
R
,
Taye
M
,
Besseling-van der Vaart
I
,
Ujčič-Voortman
J
,
Szajewska
H
,
Seidell
JC
, et al
.
The relationship of prenatal and infant antibiotic exposure with childhood overweight and obesity: a systematic review
.
J Dev Orig Health Dis
.
2020
;
11
(
4
):
335
49
. .
7.
Wan
S
,
Guo
M
,
Zhang
T
,
Chen
Q
,
Wu
M
,
Teng
F
, et al
.
Impact of exposure to antibiotics during pregnancy and infancy on childhood obesity: a systematic review and meta-analysis
.
Obesity
.
2020
;
28
(
4
):
793
802
. .
8.
Mukhopadhyay
S
,
Sengupta
S
,
Puopolo
KM
.
Challenges and opportunities for antibiotic stewardship among preterm infants
.
Arch Dis Child Fetal Neonatal Ed
.
2019
;
104
(
3
):
F327
F332
. .
9.
Flannery
DD
,
Ross
RK
,
Mukhopadhyay
S
,
Tribble
AC
,
Puopolo
KM
,
Gerber
JS
.
Temporal trends and center variation in early antibiotic use among premature infants
.
JAMA Netw Open
.
2018
;
1
(
1
):
e180164
. .
10.
Villar
J
,
Giuliani
F
,
Figueras-Aloy
J
,
Barros
F
,
Bertino
E
,
Bhutta
ZA
, et al
.
Growth of preterm infants at the time of global obesity
.
Arch Dis Child
.
2019
;
104
(
8
):
725
7
. .
11.
Wood
CT
,
Linthavong
O
,
Perrin
EM
,
Leviton
A
,
Allred
EN
,
Kuban
KCK
, et al
.
Antecedents of obesity among children born extremely preterm
.
Pediatrics
.
2018
;
142
(
5
):
e20180519
. .
12.
Li
DK
,
Chen
H
,
Ferber
J
,
Odouli
R
.
Infection and antibiotic use in infancy and risk of childhood obesity: a longitudinal birth cohort study
.
Lancet Diabetes Endocrinol
.
2017
;
5
(
1
):
18
25
. .
13.
Fenton
TR
,
Kim
JH
.
A systematic review and meta-analysis to revise the Fenton growth chart for preterm infants
.
BMC Pediatr
.
2013
;
13
:
59
. .
14.
World Health Organization
.
WHO child growth standards: length/height-for-age, weight-for-age, weight-for-length, weight-for-height and body mass index-for-age: methods and development
.
World Health Organization
;
2006
.
15.
Jobe
AH
,
Bancalari
E
.
Bronchopulmonary dysplasia
.
Am J Respir Crit Care Med
.
2001
;
163
(
7
):
1723
9
. .
16.
Hauser
RM
.
Measuring socioeconomic status in studies of child development
.
Child Dev
.
1994
;
65
(
6
):
1541
5
. .
17.
Block
JP
,
Bailey
LC
,
Gillman
MW
,
Lunsford
D
,
Daley
MF
,
Eneli
I
, et al
.
Early antibiotic exposure and weight outcomes in young children
.
Pediatrics
.
2018
;
142
(
6
):
e20180290
. .
18.
Fajardo
C
,
Alshaikh
B
,
Harabor
A
.
Prolonged use of antibiotics after birth is associated with increased morbidity in preterm infants with negative cultures
.
J Matern Fetal Neonatal Med
.
2019
;
32
(
24
):
4060
6
. .
19.
Gerber
JS
,
Bryan
M
,
Ross
RK
,
Daymont
C
,
Parks
EP
,
Localio
AR
, et al
.
Antibiotic exposure during the first 6 Months of life and weight gain during childhood
.
JAMA
.
2016
;
315
(
12
):
1258
65
. .
20.
Aris
IM
,
Lin
PID
,
Rifas-Shiman
SL
,
Bailey
LC
,
Boone-Heinonen
J
,
Eneli
IU
, et al
.
Association of early antibiotic exposure with childhood body mass index trajectory milestones
.
JAMA Netw Open
.
2021
;
4
(
7
):
e2116581
. .
21.
Pyle
AK
,
Cantey
JB
,
Brown
LS
,
Heyne
RJ
,
Wozniak
PS
,
Heyne
E
, et al
.
Antibiotic exposure and growth patterns in preterm, very low birth weight infants
.
Matern Health Neonatol Perinatol
.
2021
;
7
(
1
):
7
. .
22.
Reid
BM
,
Eisenberg
R
,
Forman
K
,
LaTuga
MS
.
Relationship between early antibiotic exposure and short-term growth velocity in premature neonates
.
Am J Perinatol
.
2019
;
36
(
10
):
1014
22
. .
23.
Rajar
P
,
Saugstad
OD
,
Berild
D
,
Dutta
A
,
Greisen
G
,
Lausten-Thomsen
U
, et al
.
Antibiotic stewardship in premature infants: a systematic review
.
Neonatology
.
2020
;
117
(
6
):
673
86
. .
24.
Martínez-Jiménez
MD
,
Gómez-García
FJ
,
Gil-Campos
M
,
Pérez-Navero
JL
.
Comorbidities in childhood associated with extrauterine growth restriction in preterm infants: a scoping review
.
Eur J Pediatr
.
2020
;
179
(
8
):
1255
65
. .
25.
Ceratto
S
,
Savino
F
,
Vannelli
S
,
De Sanctis
L
,
Giuliani
F
.
Growth assessment in preterm children from birth to preschool age
.
Nutrients
.
2020
;
12
(
7
):
1941
. .
26.
Dawson-Hahn
EE
,
Rhee
KE
.
The association between antibiotics in the first year of life and child growth trajectory
.
BMC Pediatr
.
2019
;
19
(
1
):
23
. .
27.
Kamphorst
K
,
Oosterloo
BC
,
Vlieger
AM
,
Rutten
NB
,
Bunkers
CM
,
Wit
EC
, et al
.
Antibiotic treatment in the first week of life impacts the growth trajectory in the first year of life in term infants
.
J Pediatr Gastroenterol Nutr
.
2019
;
69
(
1
):
131
6
. .
28.
Azad
MB
,
Owora
A
.
Is early-life antibiotic exposure associated with obesity in children
.
JAMA Netw Open
.
2020
;
3
(
1
):
e1919694
. .
29.
Schmidt
B
,
Roberts
RS
,
Davis
P
,
Doyle
LW
,
Barrington
KJ
,
Ohlsson
A
, et al
.
Long-term effects of caffeine therapy for apnea of prematurity
.
N Engl J Med
.
2007
;
357
(
19
):
1893
902
. .