Introduction

Esophageal atresia (EA) is a rare malformation with a mean incidence of 162 per year in Germany [1]. In spite of medical and surgical progress, many affected children suffer from life-long gastrointestinal problems such as relevant gastroesophageal reflux (GER) leading to esophagitis, esophageal motility problems, dysphagia, necessity for tube feeding, and failure to thrive. Up to 40% of patients are affected by respiratory symptoms [2]. Feeding difficulties are multifactorial, including esophageal, oropharyngeal and behavioural disorders [3]. Furthermore, esophageal atresia is associated with other congenital malformations in over half of all cases [4, 5], such as cardiac abnormalities (28 to 42% [4, 5]), which may influence growth and development [2]. Low birth weight is also frequently associated: 17 % of children with EA are born small for gestational age [5] and around 40% of children are born preterm [5, 6]. A series of 287 patients reports a birthweight below 1500 g in 14% of newborns with EA [4].

While a reduced body weight and height was described in many studies [6,7,8,9], factors of comorbidity and prematurity have not been adequately considered [2]. A recent study of the National French EA Registry including more than 700 patients identified prematurity (odds-ratio 2.4), low birth weight (small for gestational age, odds-ratio 2) and other co-morbidity as risk factors for undernutrition or stunting at the age of one year, independent of the type of EA or technique of repair [5]. Bodyweight at discharge was shown to be predictive of bodyweight-for-age at the ages of one and three years. The average weight-for-age score might be close to the general population by the age of three years [10]. Another German series reports a body weight below the 25th percentile in 67% of six-year-olds and 41% of patients at the age of ten years [11]. At the age of 12 years, the average height-for-age was − 0.2 (standard deviation 0.13). Ultimately, only 7% of patients remain below the 5th percentile for body weight and height in long-term follow-up [12]. In this case series with 371 patients by Deurloo et al., prematurity and GER were identified as risk factors for low weight and height [12].

Current standard reference percentile curves for infants and children 0 through 18 years in Germany were published in 2001 by Kromeyer-Hausschild et al. and are based on a systematic review of 17 regional studies [13]. In 2013, new percentiles based on the KiGGS cohort longitudinal study on the health of children, adolescents and young adults were published [14], that are very similar to Kromeyer-Hauschild, but provide more detailed data for infants and children below the age of two years [14]. For newborns, percentiles according to gestational week were published in 2006 by Voigt et al. [15].

Unfortunately, there is no national registry for children with EA in Germany. However, in 2012, the national patient organization for EA (Kinder und Erwachsene mit kranker Speiseröhre, KEKS e.V.) established a structured follow-up plan for their members, which is widely utilized and accepted in the pediatric surgical community. Families and healthcare-providers are thereby encouraged to transmit follow-up data to an anonymized database.

The aim of this study was to calculate specific percentiles for body weight and height for children born with esophageal atresia according to sex from birth to the age of 6 years, accounting for prematurity and presence of congenital heart disease.

Methods

Data source

Anonymized data from a voluntary registry for patients born with esophageal atresia collected in the years 2012 through 2021 by KEKS e.V. was analysed. In this paper-based registry, that provides a standardised follow-up plan for patients, standardised forms are completed by the attending physician and forwarded to the patient organisation, where they are digitalised in an anonymous data base.

Data extraction

Patients born between November 15, 2001 and November 15, 2021, were included from birth to 72 months of age. For preterm children, chronological age at the time of the examination was corrected for gestational week during the first year of life. Examinations with missing information on sex or both bodyweight and -height were excluded.

Imputation of data and statistical analysis

For examinations, missing information on either weight or height, data was imputed. Assuming data was missing at random, we used multiple imputation and fully conditional specification methods to retrieve missing information [16, 17]. For imputation purposes, we used information on age and accounted for sex. For each combination of sex and age group, 100 data sets were generated in order to represent a convenient random sample of missing values. Finally, since data of four examinations was available for each patient on average, we used a multilevel approach to account for the clustered nature of our data [18]. For statistical analysis, we grouped patients according to chronological age corrected for gestational age: at birthdate, at monthly intervals for patients aged 1 to 12 months, quarterly intervals for patients aged 13 through 24 months, and intervals of six months for patients aged 25 through 72 months of age at the index examination.

Evaluation compared to the general population

Data was compared to percentiles based on the KiGGS cohort longitudinal study on the health of children, adolescents and young adults [14]. For birth weight and birth length, z-scores were calculated according to percentiles for newborns by Voigt et al. [15]. Small for gestational age was defined as below the 10th percentile or z-score below − 1.28 for sex and gestational week. Means and 95%-confidence intervals (95%-CI) were calculated.

Evaluation of influencing factors

Sex differences and the effect of congenital heart disease were evaluated using quantile regression models [19]. Quantile regression parameters were estimated for the following percentiles: 10th, 25th, 50th, 75th, and 90th. The resampling method, which uses bootstrap procedures, was used to calculate 95%-CI.

Statistical analysis was performed using SAS software Version 9.4 (SAS Institute, Inc., Cary, NC, USA). For multiple imputation Blimp-Studio was used [20].

Results

In total, 1812 examinations of 485 patients born with EA during the study period of ten years were considered. Most data was available at birth (34% of examinations for weight and 36% of examinations for length) and the first year of life (42% of examinations for weight and 39% of examinations for length). Due to missing data on either sex, age older than 72 months or missing information on both bodyweight and height, a total of 184 patients and 580 single examinations were excluded (Fig. 1). Finally, 1232 examinations of 301 patients were included. The variable bodyweight was imputed in 40 (3.3%) examinations, and height in 154 (12.5%) examinations (Fig. 1). More detailed information on long-term medical history is provided in Table 1.

Fig. 1
figure 1

Inclusion Flowchart

Full size image
Table 1 Patient characteristics
Full size table

Gestational age, birth weight and -length

Data on gestational age, bodyweight and -length at birth were available for 276 patients (118 female, 158 male). The information on length was missing for 28 newborns. In total, 98 children (32.6%) were born premature and 112 (40.7%) small for gestational age. The mean gestational age was 37.1 weeks (95%-CI: 36.7–37.6, range 26–42). The mean body weight-for-age at birth was − 1.0 (95%-CI: − 1.1; − 0.9) and length-for-age − 0.6 (95%-CI: − 0.7; − 0.4) according to gestational age [15]. There was no difference in gestational age between sexes (male 37.1 weeks, 95%-CI: 36.5–37.7, female 37.2 weeks, 95%-CI: 36.5–37.8). Males had a slightly lower mean z-score for birth weight (male − 1.1, 95%-CI: − 1.2; − 0.9, female − 1.0, 95%-CI: − 1.1; − 0.8) and length (male − 0.6, 95%-CI: − 0.8; − 0.4, female − 0.5, 95%-CI: − 0.7; − 0.2).

Percentiles

Percentiles for bodyweight and –height for EA patients according to gender are presented in Tables 2 and 3, respectively. Children born with EA had a lower median for bodyweight and length compared to the general population for the most part (Figs. 2a, b and 3a, b). In both males and females, median bodyweight developed along the 10th percentile of the general population (Fig. 2a, 3a). The median height developed along the median of the general population during the first six months of life, before crossing percentiles to develop along the 10th of the general population around the age of 12 months (Fig. 2b, 3b). All curves of EA patients show a certain fluctuation due statistical uncertainty that must be considered while interpreting the data.

Table 2 Percentiles for bodyweight and –height in males with esophageal atresia according to age (months) after multiple imputation. Germany, 2001–2021. Source: patient registry of KEKS e.V.
Full size table
Table 3 Percentiles for bodyweight and –height in females with esophageal atresia according to age (months) after multiple imputation. Germany, 2001–2021. Source: patient registry of KEKS e.V.
Full size table
Fig. 2
figure 2

Percentiles for body weight (a) and height (b) for males with esophageal atresia (EA) during the first 6 years of life compared to the general male population [14]. Germany, 2001–2021. Source: patient registry of KEKS e.V.

Full size image
Fig. 3
figure 3

Percentiles for body weight (a) and height (b) for females with esophageal atresia (EA) during the first 6 years of life compared to the general male population [14]. Germany, 2001–2021. Source: patient registry of KEKS e.V.

Full size image

Impact of congenital heart disease

The information on presence or absence of congenital heart disease (CHD) was available for 276 patients (91.7%). At birth, there was no difference in gestational age between children with or without additional congenital heart disease (Table 4). Children with CHD had a lower z-score for birth weight and length (Table 4). During the first year of life, the presence of additional CHD has a substantially higher impact on body weight than on height. At the age of 13 to 24 months, CHD was associated with a higher median weight compared to EA patients without CHD, especially in females (Supplement 1a/ 1b). Due to sparse information in the older age groups of 13 to 24 months and more pronounced of25 to 72 months, estimates of the impact of CHD were characterized by wide confidence intervals (Supplement 1a/ 1b).

Table 4 Means and 95%-confidence intervals (95%-CI) of gestational age and biometric z-scores *according to Voigt [15] of esophageal atresia patients with and without additional congenital herat disease (CHD). Germany, 2001–2021. Source: patient registry of KEKS e.V.
Full size table

Discussion

Our study provides percentiles for caregivers and health professionals specifically for children born with esophageal atresia, which have previously not been available. Compared to the general population, patients with EA are born with a lower birth weight and -length, and thereafter tend to follow a different, specific growth pattern. Furthermore, male infants and infants with additional congenital heart disease had a lower mean z-score for body weight at birth.

During the past decades, improvements in neonatal and surgical care, as well as early nutrition interventions have reduced the risk of malnutrition [2]. In our study, median weight for children with EA develops along the tenth centile of the general population. While the median height is close to normal during the first 6 months of life, growth decelerates thereafter in the sense that height crosses percentiles between the ages of 6 and 15 months and develops along the tenth centile of the general population until the age of 6 years. This finding corroborated the results of the French EA Registry, showing no catch-up growth during the first year of life, but a harmonious growth retardation in some patients at the age of one year [5]. Feeding problems seem to have an immediate impact in body weight, but a delayed impact on body height in patients with EA. Typically, more solid foods are introduced into an infant’s diet at the age of five to six months in order to promote age-appropriate development. In EA patients, esophageal strictures that had no impact on a formula diet, may become symptomatic and impact on adequate nutrition. They therefore should be identified and treated without delay in this important and consequential developmental phase. An additional adverse factor may be that parents typically hesitate to introduce solid foods. Therefore, children require close developmental, as well as weight and height surveillance during this vulnerable phase to detect imminent problems. Dysphagia can have many underlying causes in EA patients and is a common problem with a prevalence ranging from 21 to 84% of patients of all ages [2].

About one third of our cohort was born prematurely. This fact was taken into account by correcting the chronological age for the gestational age regarding all data on body weight and length during the first year of life. The influence of additional CHD was analysed by quartile regression and demonstrated an effect that should be taken into account, especially for older children with additional CHD while regarding our percentiles.

Long-time follow-up demographic data is scarce for patients born with EA. According to a recent study, EA patients still have a lower mean z-score for body weight (− 1.4) and height (− 0.7) in adolescence around the age of 15 [21]. One study reported that 75% of 13-year-old patients had a normal nutritional status, 16% were obese and 9% were malnourished [22]. Another study found a normal nutritional status in 68% of adolescents and young adults with EA at the age of 14 through 18 years [23]. Interestingly, in patients older than 18 years, there is a near-normal distribution of percentiles for body weight and –height [7]. Further studies in the adolescent and adult population are needed to determine at what age catch-up growth and weight gain occur.

Strengths and limitations

This is the first study derived from a large national sample of patients born with esophageal atresia in Germany. Data quality is limited by the voluntary nature of data entry to the registry. While in some cases, information such as sex is missing, the overall data is qualitatively and quantitatively robust. As expected, registry entrances decrease with increasing age, leading to a larger mathematically-induced variation in the percentile curves (Figs. 2 and 3) in older children. We overcame these limitations by employing a multiple imputation approach for missing data.

It is also important to note that the growth data is cross-sectional for patients with esophageal atresia and not disease-specific. Co-morbidities may impact on growth, as well as the complexity of the individual patient’s case. Addressing the effects of different co-morbidities and other individual factors should be part of larger, future studies.

Conclusion

Children with EA had a reduced bodyweight and -height compared to the general population. The age-based introduction of solid foods is a vulnerable phase, where occult problems, such as strictures, may become symptomatic and must be addressed without delay in order to ensure timely development and growth. Cardiac co-morbidities further compromise weight gain in these patients, implying that such patients should be under even closer surveillance.