Preterm birth has become the leading cause of death and contributes significantly to long-term morbidities worldwide(Reference Walani1). Indonesia is estimated as the top five countries with the highest preterm birth rate, about 10·4 per 100 live births or nearly 550 000 preterm births each year, contributing 3·5 % of global preterm births(Reference Blencowe, Cousens and Oestergaard2,Reference Chawanpaiboon, Vogel and Moller3) . Maternal nutrition-related factors are still major risk factors for preterm birth in Indonesia(Reference Haksari4–Reference Sebayang, Dibley and Kelly7).
Advances in neonatal care management in Indonesia have improved the survival rate of preterm infants as seen in other countries(Reference Owen, Manley and Davis8–10). However, concern about suboptimal postnatal growth and development among this high-risk population remains, especially after discharge from the hospital, where growth monitoring is often either neglected or forgotten; thus, close follow-up is a necessary strategy(Reference Saigal and Doyle11,Reference Saigal12) . Previous cohorts in Asian developing countries reported a significant delay in weight, length and head circumference of preterm infants after discharge from the neonatal intensive care unit (NICU)(Reference Wang, Fang and Tsou13–Reference Japakasetr, Sirikulchayanonta and Suthutvoravut15). Growth after discharge in this population may be influenced by baseline complexity of co-morbidities during NICU stay, nutritional status at discharge, post-discharge nutrition and also socio-economic condition.
Continuous adequate nutrition and stimulation after discharge are essential not only for preterm infants’ survival but also for physical growth, cognitive and mental development and later productivity across the entire life span(Reference Ramel, Demerath and Gray16,Reference Lucas, Morley and Cole17) . Poor postnatal weight gain in preterm infants has been shown to harm neurodevelopment as physical growth during the first 1000 d of life is an important proxy for organ development, especially the brain’s maturation, myelinisation and synaptogenesis. Therefore, NICU graduates need to catch up with growth by 24 months of corrected age. However, nutrition therapy had to be handled delicately because rapid postnatal growth crossing weight centiles might lead to insulin resistance and metabolic syndromes(Reference Parkinson, Hyde and Gale18,Reference Toftlund, Halken and Agertoft19) .
There has been no data on the long-term growth trajectory of preterm infants in Indonesia. This study is the first longitudinal study in Indonesia that follows preterm infant growth after discharge. This study aimed to describe the growth trajectory of preterm infants discharged from NICU until 24 months of corrected age and its variation among gestational age groups. This work also aims to give insights and guide clinicians in the growth of preterm infants to prevent long-term consequences related to preterm birth in Indonesia.
Methods
Study design and setting
This is a longitudinal study that repeatedly measures infants’ growth parameters from NICU discharge until 24 months of corrected age. This study was conducted from January 2018 to December 2020 in the Cipto Mangunkusumo Hospital, Jakarta, a tertiary-level referral hospital. Cipto Mangunkusumo Hospital provide the most advance neonatal care and also a centre for neonatologist training centre in Indonesia. This observational study was conducted according to the guidelines laid down in the Declaration of Helsinki. This study was approved by the Research Ethic Committee of Faculty of Medicine, Universitas Indonesia, Jakarta.
Study population
Eligibility criteria were preterm infants (gestational age < 37 weeks) discharged from NICU during the study period. Exclusion criteria were infants with severe congenital anomalies or syndromes, conditions that interfere with anthropometric parameters (e.g. hydrops and hydrocephalus), large-for-gestational age (birth weight > 90th percentile), small for gestational age (birth weight < 10th percentile), intrauterine growth restriction (as reported by obstetrician) or having anomalies related to feeding difficulty (such as cleft lip or laryngomalacia) and living outside Jakarta. At the time of subject recruitment, almost 30 % infants were classified as small for gestational age or having intrauterine growth restriction.
Assessment and follow-up
The gestational age was determined by obstetric ultrasound examination. We classified preterm infants based on WHO subcategories: late preterm (34 to < 37 weeks), moderate preterm (32 to < 34 weeks), very preterm (28 to < 32 weeks) and extreme preterm (< 28 weeks). Baseline characteristic data, including neonatal sepsis, hemodynamically significant persistent ductus arteriosus, intraventricular haemorrhage and necrotising enterocolitis, were taken to describe the conditions and treatment of infants during intensive care in the NICU.
Patients were discharged if at least 36 weeks of corrected age and had appropriate weight gain. After discharge, parents were asked to visit paediatric outpatient clinic either in our hospital or the nearest hospital. We contacted parents by phone or text message regularly to remind them to visit the outpatient clinic. According to Indonesian guidelines, regular visits are recommended each month in the first 6 months and then every 3 months until 24 months of corrected age. Subjects might come earlier if asked by the paediatrician. We specifically obtained data at 0, 3, 6, 9, 12, 15, 18, 21 and 24 months of corrected age with a tolerance of ±2 weeks for each time point. Any anthropometric examination done outside our hospital was recorded according to data in the medical record as reported by parents or paediatricians.
Anthropometric examination (weight and height) was carried out by a specially trained nurse under the supervision of a physician using calibrated tools. The measurement was then plotted on the WHO Child Growth Standard weight-for-age (WAZ), height-for-age and weight-for-length (WFL) curves. Data regarding anthropometric measurements were collected by research assistants from patient medical records.
We also provide the type of nutrition at discharge and during follow-up. The type of enteral nutrition consists of breastmilk, human milk fortifier and formula (preterm formula (0·81 kcal/ml), standard formula (0·67 kcal/ml), high-density formula (1 kcal/ml) or hydrolysed or amino acid-based formula).
Outcomes
The main outcome of interest in this study is trends of WAZ and WFL and also the proportion of stunted infants (defined as height-for-age below –2 z-score according to the WHO Child Growth Standard). We also analysed the differences in trends and proportions across gestational age groups.
Data analysis
Data were extracted from an integrated online database and rearranged based on the time of clinic visit to construct a longitudinal data presentation. For standardisation, weight measurement was converted from grams and centimetre to the z-score of WAZ, height-for-age and WFL using ‘zanthro’ command in STATA. Outlier data (> 99th or < 1st centile on the growth chart) were imputed to the average values based on gestational age groups. A growth trajectory was constructed by connecting the median z-score value of each time interval with a linear line. We also provide the 95 % CI data for each timepoint. The proportion of infants who were stunted were presented as number and percentage. To optimise analysis, a subgroup merge will be conducted if a gestational age group were having lower than thirty participants.
Adjustment for potential confounders (energy intake and parental stimulation) was not necessary as nutrition and stimulation recommendations were personalised for each subject during an outpatient visit. A brief 24-h food recall is a standard procedure to evaluate the adequacy of energy intake. Any intervention for faltering growth might include nutrient intake by using high-dense caloric formula, parental education of feeding rules and any investigation for acute illness. An individual incomplete specific time interval data will be imputed according to sex and gestational age rather than excluded in order to minimise the risk of selection bias.
The difference in weight or length during the time interval was analysed using repeated measure ANOVA, the difference between two time intervals was analysed using paired t test or Wilcoxon test, while the difference in proportions was assessed using χ 2 or Fisher exact test. Analysis of trajectory trends between gestational age groups was conducted using generalised estimating equation regression. Based on the sample size calculation to estimate the prevalence of underweight, stunted and malnutrition, a minimum of 300 subjects were required to achieve a power study of 80 %. A P-value < 0·05 was considered statistically significant. Data analysis was done by using STATA v15.
Results
During the study period, 336 preterm infants were discharged from the neonatal unit, and 306 of them were eligible to participate in this study (Fig. 1). The reason for exclusion was due to living outside Java Island. Almost two-third were females, and the overall mean birth weight was 1804 (±415) g. Most of the infants were moderate preterm and had low birth weight. Causes of preterm birth were mostly due to preeclampsia and intrauterine infection.
Subject recruitment flow diagram. SGA, small for gestational age; IUGR, intrauterine growth restriction; LGA, large for gestational age; NICU, neonatal intensive care unit.
The highest prevalence of co-morbidities was neonatal sepsis, which accounted for 21·25 %. Necrotising enterocolitis was found in 11·25 % of infants. At discharge, female had slightly higher weight, length and WFL z-score. Table 1 shows the baseline characteristics of the included subjects.
Baseline characteristics of preterm infants during neonatal intensive care unit admission (Numbers and percentages; mean values and standard deviations; median values and interquartile ranges)
* Presented as median (IQR).
At the first 6 months of corrected age or before the introduction of solid foods, 70·4 % of infants received breastmilk (25·2 % exclusive breastmilk, while the rest received a combination with formula), and 32·8 % among them had an additional human milk fortification. About 53 % of infants received formula, mostly high-density calorie formula (1 kcal/ml) to support growth. The type of formula varied based on the paediatrician’s nutritional assessment.
An observation of the weight trend was presented in 3 months interval, as seen in Fig. 2. Overall WAZ at 0 months was in the median of the curve. A drop in z-score at 2–3 months (from z-score –0·16 to –1·22, P< 0·001) was followed by a gradual increase to –0·87 at 9 months and finally to 0·38 at 24 months, higher than weight at 0 months (P< 0·001).
Overall median and 95 % CI weight trajectory in weight-for-age (WAZ) from 0 to 24 months corrected age. Red line refers to reference for normal WAZ value.
We then analysed the WAZ based on gestational age groups (Fig. 3). Due to low number of extreme preterm, we merge very-to-extreme preterm into one group. All groups had a normal range of Z-score at 0 months and were found to have growth decrement in the first 3 months after birth, with the biggest decrease seen in the very-to-extreme preterm groups (decrease 2·44 ± 0·14 z-score, P< 0·001). After that, they all had an improvement in WAZ gain significantly from 3 months to 24 months. Interestingly, there were a second period of WAZ drop among very-to-extreme preterm at 6–9 months and among late preterm at 11–13 months. The late preterm and moderate preterm groups showed an earlier start of WAZ gain compared with the extreme to very preterm group. At 24 months, the WAZ of all groups were above the median of the curve: 0·28, 0·38 and 0·67 for late, moderate and very-to-extreme preterm infants, respectively, (P< 0·001). The overall different weight trajectory trends between gestational age groups were statistically NS (P= 0·263).
Median and 95 % CI weight trajectory in weight-for-age (WAZ) from 0 to 24 months corrected age for late (blue line, circle dot), moderate (green line, triangle dot) and very-to-extreme (red line, square dot) preterm infants. Red line refers to reference for normal WAZ value.
In general, almost half (49·4 %) of included infants were stunted at 3 months of corrected age. The percentage declined along with the growth of the infants, as shown in Table 2. The percentage of stunted infants (online Supplementary Fig. 1) was then decreased to 44·4 %, 23·5 %, 31·4 %, 18·9 % and 13·4 % at 6, 9, 12, 18 and 24 months, respectively (P< 0·001).
The proportion of stunted at 3, 6, 9, 12, 18 and 24 months based on gestational age groups (Numbers and percentages)
Moderate preterm had the highest proportion of stunted at the first 6 months but then declined due to catch-up growth. At 12 months, the very-to-extreme preterm group had the highest proportion of stunted. The proportion differed over time (P< 0·001) but at an equal proportion in 24 months (P= 0·205).
Using both weight and height information, we construct a graph of WFL z-score for 24 months, as displayed in Fig. 4. During the follow-up period, regardless of the WAZ and stunted status, all infants were in the normal range of WFL (between –2 and 2 z-score) although we found a decrement during the 6–12 months period. The trends between gestational age groups were visually similar but statistically differ over time (P = <0·001) as seen in online Supplementary Fig. 2.
Overall median and 95 % CI weight-for-length (WFL) trajectory from 0 to 24 months corrected age. Red line refers to reference for normal WFL value.
Discussion
This is the first longitudinal study in Indonesia to describe the growth trajectory of preterm infants after discharge until 24 months of corrected age. In sum, preterm infants had a decrement of WAZ during the first 3 months of corrected age and gained WAZ gradually until 24 months to meet the median of the curve. Almost half of them were categorised as stunted at discharge but gradually decreased at 24 months. All infants in this study had good growth trajectories as they were always in the normal range for WFL z-score. We believe this result is an important finding in neonatal care, especially for developing countries. These results also suggested that close follow-up of preterm infants after NICU discharge is crucial as growth faltering might occur anytime during the first 24 months.
Some important note is that the survival rate of very and extreme preterm in Indonesia during the study period (2018–2020) is low. Our unpublished data reported 24·9 % and 47·4 % 1-month survival rates for extreme and very preterm, respectively. Survived infants tend to have fewer co-morbidities, while procedures- and environment-related stressors during the critical phase strongly affect long-term outcomes(Reference Smith, Gutovich and Smyser20,Reference Grunau, Whitfield and Petrie-Thomas21) . This might explain the higher WAZ and WFL of very-to-extreme preterm infants at 0 months of corrected age. However, they were all regressed into the mean at 24 months along with other gestational age groups.
The major challenge in conducting a longitudinal study in developing countries after discharge was encouraging parents to make regular visits, especially after 6 months of corrected age. Good communication should be made to inform parents regarding the importance of close follow-up, even in healthy ex-preterm infants as it could bring advantages to growth and development(Reference Ballantyne, Benzies and Rosenbaum22). An optimal recommendation is a monthly visit during the first 24 months, but some reported a decreasing trend after 12 months(Reference Wade, Lorch and Bakewell-Sachs23,Reference Taslimi Taleghani, Fallahi and Soltanttooyeh24) . On the other hand, a multicentre study in Europe reported that parents’ dissatisfaction with post-discharge healthcare delivery was high(Reference Seppänen, Sauvegrain and Draper25). Otherwise, an over-visit might be a sign of parental anxiety(Reference Allen, Manuel and Legault26).
In our hospital, enteral nutrition was started from the first day of life and advanced based on feeding tolerance. Aggressive total parenteral nutrition was given to infants born under 32 weeks of gestational age. Full feed target should be reached in 1 week, 1–2 weeks or 2–3 weeks in late, moderate and very-to-extreme preterm infants, respectively. Measurable factors related to malnutrition were hospital malnutrition (postnatal growth failure) that can occur before discharge; thus, nutrition during NICU stay is as important as post-discharge feeding practice(Reference Liao, Lin and Wang27,Reference Rohsiswatmo, Kaban and Sjahrulla28) . Complementary foods were commonly given at 6 months but can be started as early as 4 months, according to the growth and readiness of the infants. The WHO had recommended that infants should be exclusively breastfed for the first 6 months in order to achieve optimum growth and development(29). The Indonesian Pediatric Society does not recommend starting complementary feeding before 4 months.
The weight trajectory was nearly a U-shaped with nadir point at 3 months and gradually increasing to 24 months. Similar to the result of this study, previous research in Minneapolis reported that weight was at the lowest point in 3–4 months corrected age (between z-score –1·0 and –1·5) and gradually increased until 24 months(Reference Ramel, Demerath and Gray16). A cohort in India also showed a similar growth trajectory to our study(Reference Mukhopadhyay, Mahajan and Louis14). Among infants < 34 weeks and < 1500 g, weight decreased from –2·3 at 0 months to –2·5 at 3 months, gradually increasing to –1·7 at 12 months. Weight z-score at 3 months was found to be an independent predictor of being underweight at 12 months (OR 4·34; 95 % CI 2·38, 7·69)(Reference Mukhopadhyay, Mahajan and Louis14). On the other hand, a population-based study in China revealed the fastest period of catch-up growth was at the first 3 months of corrected age before the z-score fell(Reference Kang, Wang and He30). Another study by Han et al. reported an increase in weight at the first 3 months (+0·08 z-score) but then decreased until 24 months among appropriate for gestational age preterm infants(Reference Han, Jiang and Huang31). During this period, a close follow-up is needed to ensure nutritional adequacy by adjusting energy intake from breastmilk or formula and also infant feeding practice by the caregiver. Half of preterm infants receive full formula feeding after 1 month of corrected age, but mostly below recommended energy intake(Reference Ribas, Cunha and Paravidino32).
The growth trajectory after 3 months, according to the change in WAZ, was a follow-the-curve growth pattern (change in z-score between –0·67 and 0·67), which is an indicator of appropriate trends(Reference Ribas, Cunha and Paravidino32,Reference Chandra and Spitzer33) . Weight gain is still useful to evaluate growth in preterm infants(Reference Greer, Norlin and Olsen34). However, rapid catch-up (change > 0·67 z-score) causes an early adiposity rebound and is related to obesity at 5 years(Reference Chandra and Spitzer33). Compared with term infants, preterm infants tend to have more body fat at term corrected age but lessened at 3 months(Reference Ramel, Gray and Ode35). It is debatable whether other parameters such as WFL or BMI will accurately present body composition in preterm infants(Reference Nagel, Desjardins and Earthman36,Reference Ferguson, Grabich and Olsen37) . In this study, all infants were in a normal range for WFL during the entire 24-month follow-up period.
Failure to regain weight in the first 3 months possibly caused the failure of length gain, as also reported previously(Reference Figueras-Aloy, Palet-Trujols and Matas-Barceló38). Stunted cases in the first months of life were higher than in 24 months of corrected age. An increase in body length is quite constant during the first 24 months(Reference Kang, Wang and He30). According to the WHO framework, stunting is a stunted case caused by impaired growth that is influenced by poor nutrition, repeated infection and inadequate psychosocial stimulation, all of which can highly occur among preterm infants(Reference Wamani, Åstrøm and Peterson39). Data on stunted vary between studies. Gonzales-Garcia et al. reported 15 % of stunted at 24 months(Reference González-García, Mantecón-Fernández and Suárez-Rodríguez40), while Mukhopadhyay et al. reported 32 % stunted at 12 months among very low birth weight infants(Reference Mukhopadhyay, Mahajan and Louis14). Extremely low birth weight infants had a significant catch of height at 3–12 months and possibly had higher height at 24 months (z-score –0·8 compared with –1·7 for very preterm)(Reference Mukhopadhyay, Mahajan and Louis14). An interesting finding in this study was that the very-to-extreme preterm group had an increased proportion of stunted at 12 months, while the other preterm group had a declining trend. Further investigation is needed to evaluate any factors that contribute to the growth trends, such as timing of complementary feeding, total duration of receiving breastmilk, caregiver feeding practice, total daily energy intake measurement and parental stimulation. A possible explanation might be a transition to full solid foods without adequate energy intake. An individualised evaluation should be made and account for parental/caregiver feeding behaviour; thus, it would be difficult to conclude a single cause of this phenomenon.
The proportion of infants receiving breastmilk in our study was quite satisfactory as it became a common problem after discharge. Exclusive breastmilk was received in 25 % of infants, higher than a previous cohort in China (10·4–19 %)(Reference Jiang and Jiang41) and similar to a study in South Carolina (27·3 %)(Reference Zhang, Donnelly and Thomas42). A Danish cohort reported that exclusive breastfeeding can reach as high as 58–68 % among preterm infants(Reference Maastrup, Hansen and Kronborg43,Reference Maastrup, Rom and Walloee44) . In our observation, the success of catch-up growth depends on total energy intake rather than the type of milk. Up to 62 % of preterm infants had insufficient energy intake (< 120 kcal/kg) after discharge(Reference Ribas, Cunha and Paravidino32). It implies that paediatricians or nutritionists had a crucial role in providing nutritional care to preterm infants after discharge. Energy intake analysis should be made upon follow-up clinic, and early intervention should be made to optimise infant growth.
Long-term outcomes of catch-up growth in preterm infants are developmental status. How large the impact of nutrition on later neurodevelopment among preterm infants is debatable. Nutrition adequacy, represented as growth, is strongly related to neurodevelopmental outcomes(Reference Cordova, Cherkerzian and Bell45). However, a recent study concluded that brain injury and low maternal education are stronger predictors of cognitive impairment instead of weight, height and head circumference at 36 weeks postmenstrual age(Reference Fenton, Nasser and Creighton46). Compared with full-term infants, preterm infants had lower cognitive performance in later years(Reference Wang, Fang and Tsou13). The younger the infants were born, the higher the risk for developmental delay(Reference Schonhaut, Armijo and Pérez47). A possible explanation is that prematurity-related morbidity (intraventricular haemorrhage, infection, necrotising enterocolitis and other co-morbidities) causes brain injury, specifically among white matter areas(Reference Rogers and Hintz48).
Limitation
We only include a small number of extreme preterm due to low survival at that time; thus, any information acquired from extreme preterm should be interpreted carefully. Insufficient data for head circumference limited the researcher to assess the prevalence of microcephaly, which can reflect growth restriction due to malnutrition and also potential neurodevelopmental problems; however, any intracranial abnormalities should be evaluated as confounding factors. Deep exploration of qualitative aspects of preterm baby care at home, such as stimulation and feeding practice, should be investigated as well.
We did not specifically analyse parental socio-demographic factors, patient family size and in-hospital factors (duration of hospitalisation, neonatal co-morbidities) as a contributing factor to long-term growth. A further exploration of this aspect is important.
We also noted that maternal nutritional factors are among the risk factors of preterm birth; however, our data are incomplete. We cannot rely on maternal BMI or mid-upper arm circumference to reflect the nutritional status. A routine laboratory examination for anaemia, infection, fatty acid profile and micronutrients such as zinc and vitamin D level is more appropriate to represent maternal nutritional status other than weight at pregnancy. Further research should be conducted to evaluate the relationship of these maternal nutritional parameters.
Further research
Future research to determine risk factors of growth and developmental delay and also sex-dependent growth trajectory and neonatal nutritional requirements in Indonesia should be done. Hence, researchers should note that anthropometric measurements should be plotted on the growth chart, and health workers should pay more attention to any deviation on the growth curve, especially in the first 3 months of corrected age. Finally, a long-term follow-up study is mandatory to evaluate the outcome of Indonesian preterm infants. Information gathered from this data analysis was then used by the authors to design a prospective cohort study (CIPTO Study) of which the protocol had been published(Reference Rohsiswatmo, Hikmahrachim and Sjahrulla49).
Conclusion
Growth trajectory trends differed between weight, length and WFL in the first 24 months. Weight reached its nadir point at 3 months and then gradually increased above the median curve at 24 months. There also was a reduction of stunted proportion at 24 months of corrected age. Otherwise, there was a normal WFL trend from discharge to 24 months. The growth trajectories were varied between different gestational age groups. Close follow-up with an individualised approach is crucial for optimal growth of preterm infants after NICU discharge.
Supplementary material
For supplementary material/s referred to in this article, please visit https://doi.org/10.1017/S0007114524002538.
Acknowledgements
The investigators thank all attending physicians and nurses at the neonatal unit of Cipto Mangunkusumo Hospital for taking care of our preterm infants, especially during the critical period.
All funding related to inpatient care and outpatient visits are provided by Indonesian national health insurance.
R. R. and H. G. H. are the primary investigators that design the study. R. R., H. G. H., D. U. N. and A. A. S. carried out the study, collect the data and conduct the data analysis. Findings are interpreted by all authors (R. R., H. G. H., D. U. N., A. T. P. I., P. M. T. M., D. S., M. A. R. S., A. K., A. A. S., R. D. R.). H. G. H. wrote the first draft and then finalised by all authors.
The authors declare that the study was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
This studies had been approved by the Ethical Committee, Faculty of Medicine, Universitas Indonesia with protocol number 19-03-0228. Written informed consent to participate in this study was provided by the participants’ legal guardian.
Data are available upon request. Further inquiries can be directed to the corresponding author.
