ABSTRACT

Objective

In this study, we aimed to evaluate growth characteristics of preterms followed up in a neonatal intensive care unit (NICU) during adolescence.

Methods

The preterms born in 2001–2002 who were followed up in the NICU of our hospital were examined in 2016. Data from the prenatal, natal, and postnatal periods, as well as at the 6–7 years, were obtained from a database with follow–up assessments during adolescence. Measures of growth (height, weight, body mass index), the percentile curves for Turkish children, and the target height formula were used in the assessment. Factors affecting growth, such as maternal problems and neonatal morbidities, were questioned.

Results

Thirty–seven preterm babies (19 females and 18 males) were included in the study. The mean gestational age was 31.5 (28–34) weeks, and the birth weight was 1298 (820–1870) grams. Six newborns (17.6%), and three adolescents (8.1%) were below 3^rd percentile for height. All cases had height percentiles within the normal range at 6–7 years of age. Twenty–six adolescents (70.3%) achieved the target height. Eleven premature (29.7%) during the neonatal period, one child (2.7%) during the school period, and one adolescent (3%) were below 3^rd percentile for weight. During the neonatal period, one case for height and two cases for weight were above 90^th percentiles. While, during school period, 6 cases for height, 2 cases for weight were above 90^th percentiles; 1 case for height, 6 cases for weight were above 90^th percentiles during adolescence. The frequency of obesity increased from 2.7% to 13.5% during adolescence. 90% of individuals below 3^rd percentile of birth weight reached normal weight percentiles during adolescence. Prenatal, natal, and postnatal risk factors, as well as duration of breastfeeding, were not found to be effective in achieving the target height during school–age and adolescent periods.

Conclusion

In this study, we found that all cases had normal percentiles for height in the school period, but on the other hand, 8.1% cases were <3^rd percentiles during adolescence. Also, the frequency of cases that were <3^rd percentiles for weight was similar during the school period and adolescence. These results reflected that, while catch–up growth could be achieved during the school years, it was not achievable during adolescence. It demonstrates that the growth of individuals born preterm is a dynamic phenomenon.

INTRODUCTION

In parallel with advances in perinatal and neonatal care worldwide, the survival of premature babies has increased, and more are being discharged from the hospital. The follow–up of high–risk preterm infants has highlighted several key questions concerning growth patterns and developmental milestones. Prematurity is associated with a range of short– and long–term complications, including impaired somatic growth, neurodevelopmental delays, and increased risk of chronic health conditions. These outcomes may persist into adolescence and adulthood, underscoring the importance of longitudinal monitoring. A holistic examination of the long–term consequences of premature birth is of great importance for prognosis.

Preterm infants are prone to early postnatal growth restriction; nevertheless, most subsequently achieve somatic catch–up growth. Increases in weight and length typically begin within the first months of life and often normalize by two years of age (1). Some postnatal comorbidities, such as bronchopulmonary dysplasia (BPD), retinopathy of prematurity (ROP), and neurodevelopmental impairment, are also reported to affect catch–up growth (2).

Numerous studies have examined long–term monitoring in very–low–birth–weight premature infants (3-5). Although studies on somatic growth and neurodevelopmental evaluation are available in infancy, early childhood, and school–age, data on adolescent follow–up are limited (6-8).

For this reason, we aimed to evaluate the somatic growth of preterm infants born in 2001–2002 who were followed up in our neonatal intensive care unit (NICU) during adolescence, with a specific focus on their anthropometric outcomes.

MATERIALS AND METHODS

Premature infants born in 2001–2002 at 28–34 weeks of gestational age (GA) according to the Ballard Score and followed up in the NICU of Ministry of Health, Bakırköy Obstetrics and Pediatrics Training and Research Hospital were included in the study. Individuals who are deceased, have metabolic disease, and have anomalies are excluded. Growth curves by Kurtoğlu et al. (9) were used to evaluate somatic growth during the neonatal period. Cases whose birth weight were <3^rd percentile (rather than the more commonly used <10^th percentile threshold) were recorded as small for GA (SGA) and divided into symmetrical or asymmetrical SGA categories. This stricter definition was chosen to increase specificity, ensuring that infants classified as SGA truly represented those at highest risk of growth failure and adverse outcomes, and to enable a more accurate comparison with cases presenting growth retardation during childhood and adolescence. Cases at the 3^rd–90^th percentile were considered as appropriate for GA (AGA), and cases >90^th percentile were considered as large for GA (LGA). Also, we categorised cases according to GA [extremely preterm (less than 28 weeks), very preterm (28 to less than 32 weeks), moderate to late preterm (32 to 37 weeks)] and birthweight (<1000 g: extremely low birthweight, 1000 to 1499 g: very low birthweight, 1500 to 2500 g: low birthweight).

The following prenatal risk factors were assessed: premature rupture of membranes, chorioamnionitis, maternal diabetes mellitus, hypertensive disorders of pregnancy, presence of chronic disease, and intrauterine growth retardation. Among natal risk factors, the following were investigated: history of fetal distress, placenta previa, and placental detachment. The following postnatal risk factors were investigated and recorded: duration of stay in the NICU, mechanical ventilation (administration and duration), surfactant therapy, sepsis, hypoglycemia, hyperbilirubinemia, respiratory distress syndrome (RDS), BPD, intraventricular hemorrhage, posthemorrhagic hydrocephalus, and ROP. Additionally, breastfeeding duration was ascertained and recorded.

The somatic growth of these children was previously evaluated at school age (6–7 years). Data from the prenatal, natal, and postnatal periods and at 6–7 years of age were obtained from the database. Values >90^th percentile during the neonatal, childhood, and adolescent periods were used as a criteria for advanced growth (10).

A single physician (KB) assessed the cohort at 14–15 years of age, following signed parental and child consent. The systemic examination of the patients was performed. Puberty status was assessed according to Tanner staging (11, 12). Ages at menarche and thelarche were recorded based on parental recall.

The standing height and weight of the cases and parents were measured by the same physician (using a Sinbo weighing instrument and a Harpenden stadiometer) at least twice, and the averages were used for analysis. Values of the cases were placed in the percentile curves prepared for the Turkish children by Neyzi et al. (13). The target length based on the mother’s and father’s heights was also calculated and checked after the percentile chart was marked. When calculating the target height, the difference between men and women, according to each society’s standards, was considered. For Turkish society, this difference was 13 cm (13, 14).

Body mass index (BMI) was calculated as weight (kg)/height (m²). Cases were identified according to the World Health Organization BMI classification (15). BMI percentiles were assessed using reference values for Turkish children (16).

This study was approved by the Ethics Committee of Ministry of Health İstanbul Kanuni Sultan Süleyman Training and Research Hospital (approval number: 15929, date: 30.09.2015). Patients and their parents provided the written informed consent.

Statistical Analysis

The findings were analyzed statistically using the Statistical Program in Social Sciences version 24.0 (2016). When evaluating the study data, McNemar test, Fisher exact test, Mann–Whitney U test, and Kruskal–Wallis test were used in addition to descriptive statistical methods (mean, median, standard deviation, frequency). Categorical variables were presented as numbers and percentages. Continuous variables were presented as mean ± standard deviation or median (minimum-maximum), depending on the distribution. The results were evaluated using a 95% confidence interval, with significance set at p < 0.05. Missing data were taken into account during analyses.

RESULTS

In this study, 37 adolescents born preterm (19 females, 18 males) were evaluated. Most cases (n = 20, 54%) were SGA; 40.5% (n = 15) were AGA; and most cases (n = 30, 80%) had birth weights of 1000–1500 grams. While the most common prenatal risk factors were intrauterine growth retardation (IUGR) and maternal hypertension, the most common postnatal risk factor was hyperbilirubinemia. Demographic characteristics and prenatal, natal, and postnatal risk factors of the cases are summarized in Table 1.

Growth in the cases during the newborn and school periods (6–7 years) was compared with adolescent growth (Table 2). During adolescence, the mean age of cases was 14.6 ± 0.78 years, the mean height was 162.3 ± 7.1 cm, the mean weight was 56.1 ± 13.8 kg, and the mean BMI was 21 ± 4.04 kg/m². Values for height, weight, BMI, and pubarche onset by gender are presented in Table 3. Age at pubarch onset was 12.06 ± 1.2 years, and at menarche onset was 12.2 ± 1.08 years. Only one case was treated for precocious puberty; in the other cases, the onset of pubertal growth and development was prompt.

While 6 cases (17.6%) were found to be <3^rd percentiles for height in the newborn period, there were no cases <3^rd percentiles for height in the school period. Three adolescents (8.1%) were <3^rd percentiles for height (Table 2). Twenty–six cases (70.3%) reached target height by adolescence (Table 4). Of the eleven cases (6 males, 5 females) that did not reach target height, 4 were AGA, 3 were symmetrical SGA, and 4 were asymmetric SGA in the neonatal period (Table 4). Eleven premature (29.7%) during the neonatal period, one child (2.7%) during the school period, and one adolescent (3%) were <3^rd percentile for weight.

During the neonatal period, one case for height and two cases for weight were >90^th percentiles. While during school period, 6 cases for height and 2 cases for weight were >90^th percentiles; 1 case for height and 6 cases for weight were >90^th percentiles during adolescence (Table 2). The Frequency of LGA was 5.4% during the neonatal period. Obesity prevalence, when evaluated according to BMI percentiles, was 2.7% during the school period and increased to 13.5% during adolescence.

To evaluate growth trajectories, height and weight percentiles at birth, at age 6–7 years, and during adolescence were examined. No statistically significant difference was observed between height percentiles at birth and in adolescence. Two of the six patients whose birth height was <3^rd percentile (33.3%) were <3^rd percentile for height currently, and the remainder (n = 4, 66.7%) reached the target height (Table 5). While all individuals with a birth weight <3^rd percentiles reached normal weight percentiles at age 6–7 (p = 0.021), 90.9% reached normal weight percentiles during puberty (p = 0.002). Intrauterine growth characteristics (SGA, AGA, LGA), parental heights, and BMI did not significantly affect this group’s attainment of the target height (Table 5).

Eight of 11 patients whose birth weight was <3^rd percentile, and 18 of remaning 26 patients whose birth weight was >3^rd percentile could reach target height. Birth weight was not found to have a statistically significant effect on target height. (p = 1.00)

When the relationship between birth height percentiles and prenatal, natal, and postnatal risk factors were examined, the incidence of hypoglycemia was higher in patients whose height <3^rd percentile (p = 0.030). No statistically significant differences were observed between other risk factors and height percentiles at birth (Table 6). There was also no statistically significant association between birth weight percentiles and the aforementioned risk factors (Table 7).

Height and weight percentiles during school age and adolescence were examined according to the intrauterine growth status of the patients (AGA, LGA, SGA), and no statistically significant differences were found. Also, no statistically significant association was found between prenatal, natal, and postnatal risk factors and height and weight percentiles during school age and adolescence.

Breastfeeding duration ranged from 15 days to 2.5 years among the cases. There was no statistically significant relationship between breastfeeding duration and patients’ growth (length and weight percentiles) during school–age and adolescence.

DISCUSSION

This study examined the somatic growth features of high–risk preterm infants, most of whom weighed <1500 grams, at approximately 14 years of age. Our results suggested that the growth trajectories of children born preterm differed across follow–up periods. Growth of the cases during the newborn and school periods (6–7 years) was compared with adolescent growth.

While 6 cases (17.6%) for height, 11 cases (29.7%) for weight were found to be <3^rd percentiles in the newborn period, all cases caught up in length gain during school age. Only one child was found to be <3^rd percentiles for weight. Three adolescents (8.1%) for height, 1 adolescent (3%) for weight were found to be <3^rd percentiles. Twenty–six cases (70.3%) could reach target height by adolescence.

Different views exist in the literature regarding catch–up growth. Catch–up growth occurs during the first three years of life in cases without severe neurological impairment, cardiac disease, or genetic disease or syndrome (17). Niklasson et al. (18) showed that normal growth could be achieved at the age of 2 years with optimal postnatal care, even in cases with IUGR, in the absence of perinatal complications and severe congenital anomalies.

Unlike our study, in a study that evaluated neurodevelopmental and physical growth of very low birth weight preterms at an average of 87 months after birth, the rate of cases <3^rd percentile was reported as 6% in height and 7% in weight (7). A study of 4423 late preterm children evaluated from birth to 14 years reported that preterm children had lower height and weight than full–term children across all age groups (19). In the Danish birth cohort in which children born between 1996 and 2003 were followed until age 18, GA was associated with height in infancy, but the difference in height between preterm and term children decreased during childhood. It has been reported that many individuals born preterm remain shorter than those born at term during adolescence. It has also been reported that BMI, which was relatively lower in preterm infants, became equal to that of term infants during childhood, with no significant difference by adolescence (20). We found the BMI during adolescence to be similar to normal values according to the WHO classification.

In the EPICure study, which evaluated extremely premature infants, the authors reported that some catch–up growth in height and weight occurred between 2.5 and 6 years of age, but growth parameters at age 6 remained below population norms. In this study, individuals born preterm were lighter and shorter at 19 years of age than the control group (21).

In a study that examined growth characteristics of very low–weight premature babies at an average postnatal age of 36 months, it was found that gender and AGA or SGA status had no effect on growth. Mechanical ventilation, chronic disease, advanced intracranial hemorrhage, ROP, and lack of breastfeeding were identified as risk factors for growth retardation (6). Akar et al. (8) emphasized in their study evaluating somatic growth in late preterm preschoolers that being AGA or SGA was not a statistically significant predictor of reaching the target height. By contrast, in a Brazilian birth cohort study, preterm children with SGA and low birth weight were shorter and thinner than term children at 5 years (22). Similar to the previous studies (6, 8), in our study intrauterine growth characteristics and the aforementioned postnatal risk factors were not associated with growth at school age or during adolescence.

Prenatal risk factors that affect growth include maternal hypertension, chronic disease, gestational diabetes, maternal smoking, and placental dysfunction (23, 24). However, our study revealed that prenatal and natal risk factors were not associated with growth at birth, during school age, and during adolescence.

Breastfeeding is also important for growth. In a study, the growth of SGA babies fed breast milk was faster than that of babies fed formula (25). In our study, no statistically significant relationship was observed between breastfeeding duration and patients’ growth, measured by height and weight percentiles, during school age and adolescence.

Our study revealed that during the neonatal and school–age periods 2 cases, during adolescence 6 cases for weight were >90^th percentiles. Two cases who were >90^th percentiles for weight at school age were asymmetrical SGA at birth. Their birth weights were 1100 grams and 1600 grams. While one case had a history of RDS, BPD, PDA, and mechanical ventilation, the other case had only hyperbilirubinemia. Of the six cases who were >90^th percentiles for weight during adolescence, 5 were SGA (asymmetric: 3, symmetric: 2) and, one was AGA at birth. Birth weights ranged from 1000 to 1600 grams. In most cases, sepsis, apnea, and hyperbilirubinemia were present as postnatal risk factors. The frequency of obesity increased from 2.7% to 13.5% during adolescence. There was no significant association of intrauterine growth status or prenatal, natal, and postnatal risk factors with weight and height in adolescence.

Few studies have examined the potential link between prematurity and the risk of childhood obesity and have yielded different results. Unlike our study, a study examining extremely preterm infants reported that females with high gains in height/weight or weight during the 2 years after NICU discharge had greater odds of obesity at 10 years, but not at 15 years (26). Another study revealed that premature infants had a greater likelihood of childhood obesity at ages 8 to 11, and no significant difference in childhood obesity risk was found between SGA and AGA preterm infants (27).

In a French cohort that prospectively evaluated very preterm infants, the obesity rate in adolescence was 13.9% (10.5–18.3%), which is similar to that was observed in our study. This study also revealed that the change in length during the neonatal hospital stay of preterm infants was negatively associated with risk of overweight or obesity at 5 and 15 years, and that the change in BMI between discharge and 2 years was positively associated that risk (28). Unfortunately, we did not have data on patients’ growth in length during their NICU stay. Another study evaluating four birth cohorts during adolescence revealed that BMI was similar between preterm and term peers, while very preterm individuals had an increased risk of overweight (29).

Growth in height accelerates during puberty, and the peak of this acceleration occurs in early puberty (Tanner stage 2) in girls and in mid–puberty (Tanner stage 4) in boys. Although the acceleration in height growth occurs earlier in girls, it ends earlier than it does in boys (30). The literature is inconsistent regarding the effect of preterm birth on timing of puberty and pubertal growth patterns. While one study assessing 129 preterms compared to terms found no differences in pubertal growth spurt (31), in another study the onset of puberty (assessed according to Tanner staging) was later for preterm girls than term controls (32). In the ESTER Preterm Birth Study, timing of pubertal growth, age at menarche or voice break were found to be similar in preterm and term individuals (33). When we evaluated our cases with respect to pubertal development, we observed that only 1 case was treated for precocious puberty, whereas the remaining cases had normal pubertal onset and development.

CONCLUSION

Although our study has certain limitations, we believe it is valuable as it includes long-term data from our country. Key limitations of our study include the participation of high-risk preterm infants requiring intensive care and the relatively small sample size, which restrict the generalizability of the findings. Multicenter studies with larger sample sizes will contribute to the literature. A further limitation of our study is the definition of SGA restricted to neonates with birth weight below the 3^rd percentile. While this approach identifies the most severely growth-restricted infants, it excludes those between the 3^rd and 10^th percentiles who are commonly included in the conventional definition. As a result, our findings may not be fully generalizable to the broader SGA population and should be interpreted with caution when compared to studies employing the standard <10^th percentile threshold. Moreover, assessing the age of thelarche based solely on parental recall constitutes a limitation in the evaluation of pubertal development.

Our findings indicate that somatic growth in preterm infants is a dynamic, non–linear process with distinct patterns across developmental stages. Although catch–up growth in height and weight was largely achieved by school age, a subset of adolescents exhibited growth failure or excessive weight gain during puberty.

These observations underscore the clinical importance of long–term, structured follow–up for preterm infants that extends beyond early childhood. Regular anthropometric monitoring into adolescence and beyond is essential to detect deviations from expected growth trajectories. Such follow–up enables timely nutritional, endocrinological, and psychosocial interventions, which may improve long–term health outcomes and quality of life.