Growth of extremely low birth weight infants at a tertiary hospital in a middle-income country

This was a pilot study to review the growth of ELBW infants at CMJAH from birth to 36 weeks postmenstrual age, audit their macronutrient intake and explore the association of prematurity complications with growth.

The study was a retrospective longitudinal description of growth and macronutrient intake in ELBW infants at CMJAH.

The study was a secondary analysis of an existing database managed using Research Electronic Data Capture (REDCAP), hosted by the University of Witwatersrand [13] and a review of relevant hospital records. Weight was captured weekly until discharge. The volume of feeds, intravenous (IV) fluids and ITN was captured daily for the first 28 days of life. The number of days to regain birth weight was captured and the mean weight velocity was calculated from the day birth weight was regained to 36 weeks postmenstrual age. Mean weight velocity was calculated using the exponential method [14] as shown below;

The Fenton’s Growth Chart [15] and Intergrowth 21st Project growth standards [16, 17] were used as reference for normal growth. For the Intergrowth 21st Project growth standards, the Intergrowth 21st Project newborn charts were used to derive the birth weight Z-scores and the Intergrowth 21st Project postnatal growth charts were used to derive the Z-scores at 36 weeks. An infant born with a birth weight below the 10th centile was referred to as small for gestational age (SGA) [18]. The delta Z-score from birth to 36 weeks postmenstrual age was determined for each infant. To avoid under-representation of infants with good growth, infants discharged before 36 weeks postmenstrual age had their discharge weight Z-score analyzed as that at 36 weeks. Delta Z-score was calculated using the formula below;

For the purpose of computation of macronutrients, the nutritional content of the various types of feeds, IV fluids and ITN was derived from product inserts and previous studies (Table 1) [19]. Daily caloric intake in kcal/kg/day was calculated for each infant during the first 28 days for life using the formula below, where volume is in milliliters (mls), weight in grams and the figure 1000 converts grams to kilograms;

Mean weekly caloric intake for each week and mean caloric intake over 28 days was calculated. A mean caloric intake of at least 110 kcal/kg/day was classified as good caloric intake. Daily protein and lipid intake were calculated using a similar formula to the daily caloric intake.

Definitions of complications of prematurity were based on those used in the Vermont Oxford Network (www.​vtoxford.​org). The need for ventilation (invasive and non-invasive), as well as complications of prematurity including NEC, retinopathy of prematurity (ROP), late onset sepsis (> 72 h after birth), oxygen on day 28 of life and patent ductus arteriosus (PDA) were recorded. Necrotizing enterocolitis and PDA were defined in this study as follows;

The unit’s feeding protocol was as follows: Prescription of feeds was at the discretion of the attending physician. Although exclusive breastfeeding was unit protocol, no donor breast milk was available. Hence, the infant formula Prenan (Nestle) was provided for those ELBW infants whose mothers were unwilling or unable to breastfeed. Newly born infants were started on intravenous fluids (usually Non-K Neonatolyte) at 80 to 100mls/kg/day on the first day of life. Feeds were introduced on the second day of life starting at 20mls/kg/day. Feeds were gradually increased by 20 to 30mls/kg/day replacing intravenous fluids until 160 to180mls/kg/day of full enteral feeds was reached. Upon reaching full enteral feeds, each feed of breastmilk was fortified with 1 g FM85 (Nestle). Enteral feeds were discontinued if there was evidence of feeding intolerance. Infants who were not on full feeds for more than 48 h due to any reason had ITN prescribed for them provided no contraindications to ITN administration were present. Infants were weighed twice a week on an electronic scale (NAGATA BW-20, Taiwan). In the absence of ongoing complications requiring hospital care, infants were discharged upon attaining 1600 g. Infants with respiratory distress syndrome (RDS) were offered nasal continuous positive airway pressure (NCPAP) and surfactant. Upon delivery, infants with RDS were transferred to a Transitional Unit where they were offered NCPAP if available. This was done while the infant awaited transfer to their final destination, the Premature Unit, where NCPAP was usually available.

Data was entered into a Microsoft Excel Spreadsheet for data cleaning. The final dataset was exported into IBM SPSS 25 for analysis. The distribution of continuous variables was explored. Those variables with a normal distribution were described using mean and standard deviation (SD) while those with a skewed distribution were described using median and interquartile range (IQR). Categorical variables were described by frequencies and percentages.

An infant with good growth was defined as that infant who regained birth weight in 21 days or less and had a weight velocity of at least 15 g/kg/day [23, 24]. An infant with poor growth was defined as that infant who regained birth weight after 21 days and/or had a growth velocity less than 15 g/kg/day. The association of growth with complications of prematurity was assessed in three ways as follows;

For continuous variables, student t-test was used for variables with normal distribution and Mann Whitney U-test for ranked data or those variables with a skewed distribution. The Chi-squared test of independence was used for categorical variables.

There were 2829 infants with birth weight less than 1500 g infants captured on the database. Of these, 886 were ELBW infants. After excluding those outside the study period, 325 infants remained. Ten infants were transferred to other hospitals and 181 infants died in hospital. Therefore, 134 infants were eligible for the study. Forty-two infants had missing records leaving a final sample of 92. For derivation of weight Z-scores and centiles, 89 were included because 3 infants had a gestational age less than 24 weeks and therefore could not be plotted on the Intergrowth 21st Project newborn charts. Characteristics of the sampled infants are displayed in Table 2.

The mean birth weight was 867 g (SD = 81.4) with a mean birth weight Z-score of −0.45 (SD = 1.01) and − 0.72 (SD = 1.20) as per Fenton Growth Chart and Intergrowth 21st Project newborn weight charts respectively. Table 3 displays the mean birth weight Z-score, weight Z-score at 36 weeks and delta Z-score as per the Fenton Growth Chart and Intergrowth 21st Project growth standard. The mean percentage weight loss was 7.38% (SD = 5.82) and the mean number of days to regain birth weight was 18.2 days (SD = 10.2). Mean weight velocity was 13.5 g/kg/day (SD = 3.1) for the full length of hospital stay. At 36 weeks postmenstrual age, the mean weight for males was 1494 g (SD = 297) with a mean weight Z-score of − 2.81 (SD = 1.16) and − 3.18 (SD = 1.84) as per Fenton Growth Chart and Intergrowth 21st Project growth standard respectively. Among females, at 36 weeks the mean weight was 1569 g (SD = 201) with a mean weight Z -score of − 2.21 (SD = 1.02) and − 1.95 (SD = 1.36) as per Fenton Growth Chart and Intergrowth 21st Project growth standard respectively. An independent samples t-test showed a significant difference in mean weight Z-scores between birth and 36 weeks postmenstrual age, p < 0.001. This was true for both reference growth charts. Therefore, infants were significantly lighter for postmenstrual age at 36 weeks compared to the time of birth. Table 4 shows how growth parameters varied across different birth weight categories.

Table 3 shows that a statistically significant increase in infants labeled as small for gestational age was observed with the Intergrowth 21st Project growth standard compared to the Fenton Growth Chart. There was a larger number of infants who plotted above the 10th centile on the Intergrowth 21st Project standards compared to the Fenton Growth Chart, but this was not statistically significant. There was no statistically significant difference between the Z-scores derived from Fenton Growth Chart and Intergrowth 21st standard for birth weight, weight at 36 weeks and delta Z score.

Figures 1 and 2 shows how the mean weekly weight for females and males respectively markedly differed from both the Fenton Growth Chart and Intergrowth 21st Project growth standards. Many infants who regained birth weight in 21 days or less did not have an adequate weight velocity afterwards and the vice versa is true as shown in Table 5. Only 25 infants (27.2%) had good growth (regained their birth weight in 21 days or less and subsequently had a weight velocity above 15 g/kg/day) (Table 5).

The mean number of days to reach full enteral feeds was 15.9 days (SD = 6.3) and by the seventh day of life, enteral feeds contributed more than 50% of caloric intake. A plateau of 90–92% mean enteral caloric contribution was reached from the twentieth day. During the first 28 days of life, the mean number of days in which infants received at least 160mls/kg/day of enteral feeds was 12.1 days (SD = 6.3). Thirty-four infants had an adequate caloric intake (above 110 kcal/kg/day). Mean caloric intake during the first 28 days was 97.0 kcal/kg/day (SD = 14.4). A mean caloric intake above 80 kcal/kg/day was reached on day 8, by day 14 the mean caloric intake was above 110 kcal/kg. There was a steady increase in mean weekly caloric intake from 65,7 kcal/kg/day (SD = 18.7) in the first, 101 kcal/kg/day (SD = 8.6) in the second, 118 kcal/kg/day (SD = 5.0) in the third to 123 kcal/kg/day (SD = 2.9) in the fourth week. The mean protein intake was 2.5 g/kg/day (SD = 0.7) with a mean daily intake above 3.0 g/kg/day being reached on the twentieth day of life. The mean lipid intake was 4.4 g/kg/day (SD = 1.1) with a mean daily intake above 5 g/kg/day being reached at the fourteenth day of life.

Pearson correlational analysis showed that during the first 28 days of life, weight velocity increased with a shorter duration to attain full feeds (r = −0.38) and the longer the infant remained on feeds of at least 160mls/kg/day (r = 0.44). Increases in protein (r = 0.38), lipid (r = 0.42) and caloric intake (r = 0.38) were associated with an increase in weight velocity during the first 28 days of life.

A comparison was done between the group classified as “good growth” (n = 25) and that classified as “poor growth” (n = 67). Complications of prematurity which showed a significant association with poor growth were CPAP for more than 2 days, invasive ventilation, oxygen on day 28 of life, late-onset sepsis, PDA and being kept nil per os (excluding the first day of life) (Table 6). No single complication of prematurity was independently associated with neither a weight above the 10th centile at 36 weeks nor a significant difference of delta Z-score between birth and 36 weeks when the Fenton Growth Chart was used. However, use of the Intergrowth 21st Project growth standards showed a significant difference in delta Z-score between infants who were receiving oxygen by day 28 (p = 0.024), those with PDA (p = 0.001) and sepsis after day 3 (p = 0.050) compared to those who did not have such complications. Being kept NPO (p = 0.123), receiving for CPAP more than 2 days (p = 0.069), antenatal steroids (p = 0.219) and invasive ventilation (p = 0.254) did not show a significant difference in delta Z-scores compared to those who did not have such complications. A Chi squared test showed a significant association between infants having a weight at 36 weeks less than the 10th centile and those who had sepsis after day 3 of life when the Intergrowth 21st Project growth standards were used (p = 0.021).