Background
Sodium is required to maintain extracellular tonicity and positive sodium balance is a prerequisite for growth.
Fluid and electrolyte management in very low-birthweight infants (VLBWI) birthweight <1500 g is critical in the first week of life and needs to be managed appropriately to prevent morbidity and mortality. Whether plasma sodium increases or decreases depends on the magnitude and direction of change in total body sodium and water. This information is often difficult to obtain in preterm infants, the cause is not always evident and incorrect treatment may be prescribed.(1,2) Sodium balance is determined by the intake or loss of both sodium and free water. The sodium requirement in VLBWI is 3–5 mmol/kg of sodium per day and their fluid requirements may go up to 200 ml/kg/day.(3) This is affected by weight, degree of prematurity and size for gestational age.(3–5)
Hypernatraemia is defined as plasma sodium level of >145 mmol/l.(1,6) In VLBWI, insensible water loss (IWL) and renal immaturity are the major contributing factors to hypernatraemia.(7) Factors affecting IWL in VLBWI are increased respiratory rate, surgical malformations (gastroschisis and omphalocoele), increased body temperature, use of radiant warmer and phototherapy.(6) Hypernatraemia is associated with an increased incidence of chronic lung disease (CLD), patent ductus arteriosus (PDA), necrotizing enterocolitis (NEC) (8) and intraventricular haemorrhage (IVH).(7) Very severe hypernatraemia is associated with poor neurological outcome.(8) It has thus been recommended that more emphasis on fluid and electrolyte therapy in VLBWI should be placed on the prevention of excessive insensible water loss rather than the replacement of insensible water loss.(9)
Hyponatraemia is defined as plasma sodium <130 mmol/l.(6,8) Water retention may be a more important cause of hyponatraemia in the VLBWI than sodium losses.(1,10) Hyponatraemia may be due to excessive fluid and water/salt administration (iatrogenic) or due to impaired tubular sodium reabsorption.(11,12) Hyponatraemia is associated with delayed growth, poor neurodevelopmental outcome,(13,14) increased risk of PDA, bronchopulmonary dysplasia, IVH, NEC (15) and neurosensory hearing loss.(16)
There is scant data regarding sodium balance in VLBWI in South Africa. Thus, the present study aimed to review sodium imbalance in VLBWI seen in Charlotte Maxeke Johannesburg Academic Hospital (CMJAH) in Johannesburg, South Africa.
Subjects and methods
The study was a retrospective cross-sectional descriptive study evaluating sodium abnormalities in VLBWI in the first week of life (birthweight between 500 g and 1500 g). All VLBWI born at Charlotte Maxeke Johannesburg Academic Hospital (CMJAH) between 1 January 2013 and 31 July 2013 were eligible for inclusion. Infants with VLBW who were admitted after 48 h of life or who died at birth and those with major congenital abnormalities were excluded from the study.
The primary objective of the study was to determine the incidence of both hyponatraemia and hypernatraemia in VLBWI at CMJAH. The secondary objective was to identify complications and mortality associated with these sodium abnormalities.
Data was collected from a neonatal computer database kept for the purpose of clinical audit. Clinical and demographic data was collected on discharge for each VLBWI by the attending medical staff. The data was verified against hospital records. Data was managed using Research Electronic Data Capture (REDCAP) hosted by the University of the Witwatersrand.(17) A copy of the laboratory result sheet is kept with each neonatal summary.
The plasma sodium level results for the first 7 days of life were obtained for each study infant from the laboratory result sheet. All VLBWIs were cared for according to unit protocols and blood was drawn for sodium measurement at the discretion of the attending physician. There was no information available on fluid intake, type of incubator (e.g. radiant warmer), cardiovascular status or whether the baby was covered in plastic. A plasma sodium level >145 mmol/l was defined as hypernatraemia and plasma sodium level <130 mmol/l was defined as hyponatraemia. Data collected for each infant studied included gestational age, birthweight, maternal steroid administration, admission temperature, phototherapy, IVH, use of continuous positive airways pressure (CPAP), mechanical ventilation, presence of PDA and NEC. Grading of IVH was according to Papile.(18) The presence of NEC was defined as either grade two or three using modified Bell's classification.(19) Outcomes were recorded as either death or survival/hospital discharge.
Statistical analysis
Data was described using standard statistical methods. Continuous variables were described using mean with standard deviation (SD) or median with range, depending on the distribution of the data. Categorical variables were described using frequencies and percentages. Comparison between different groups (survivors vs. non-survivors and hyponatraemia vs. hypernatraemia) was done using Chi-square or Fisher's exact test for categorical variables. Unpaired t-tests or non-parametric tests were used to compare continuous variables as appropriate. Only valid cases were analysed (i.e. missing information was excluded for each variable). Analysis was done using SPSS version 24 (IBM, USA).
Results
There were a total of 298 VLBWI admitted before 48 h of life, during the study period. Six infants who died before blood was drawn for urea and electrolytes were excluded from the study. There were thus 292 VLBWI in the final sample. A total of 85 VLBWI (29.01%) had sodium abnormalities and were included in the study (see Fig 1). A total of 79 patients (27.01%) had hypernatraemia while 6 patients (2%) had hyponatraemia. None of the patients had congenital abnormalities. A total of 207 of VLBWI had normal serum sodium levels in the first week of life.
Neonatal characteristics
The birthweight considered in this study was for babies born with the birthweight of 500–1500 g. The mean birthweight was 1058.5 g (SD 293.7) and mean gestational age was 28.7 weeks (SD 2.5). Other demographic and clinical characteristics are shown in Table 1. Overall survival was 64.7%. The majority of VLBWI were treated with phototherapy and nasal CPAP.
Demographics and clinical characteristics of VLBWI with abnormal sodium levels.
| Variable* | Total |
|---|---|
| Gender | 47/85 (55.3) |
| Male, n/N (%) | 38/85 (44.7) |
| Female, n/N (%) | |
| Place of birth | 72/85 (84.7) |
| Inborn, n/N (%) | 13/85 (15.3) |
| Outborn, n/N (%) | |
| Initial temp. mean (SD) | 35.5°C (1.5) |
| Steroids n/N (%), | 53/84 (62.4) |
| NVD, n/N (%), | 44/83 (51.8) |
| CPAP, n/N (%) | 62/85 (72.9) |
| Phototherapy, n/N (%) | 56/85 (72.9) |
| SVT, n/N (%) | 61/85 (71.8) |
| Outcome | 55/85 (64.7) |
| Survived n/N (%) | 30/85 (35.3) |
| Died n/N (%) | |
| PDA, n/N (%) | 8/85 (9.4) |
| NEC Stage 2/3, n/N (%) | 5/85 (5.9) |
| IVH grade 1, n/N (%) | 6/85 (7.1) |
| grade 2, n/N (%) | 7/85 (8.2) |
| grade 3, n/N (%) | 2/85 (2.4) |
| grade 4, n/N (%) | 1/85 (1.2) |
| HMD, n/N (%) | 77/85 (90.6) |
| ROP, n/N (%) | 6/85 (7.1) |
| Sepsis, n/N (%) | 2/85 (2.4) |
| Metabolic acidosis n/N (%) | 4/85 (4.7) |
| Hyperglycaemia n/N (%) | 23/85(27.1) |
*PDA = patent ductus arteriosus, NEC = necrotizing enterocolitis, IVH = intra-ventricular haemorrhage, HMD = hyaline membrane disease, ROP = retinopathy of prematurity. NVD = normal vaginal delivery, CPAP = continuous positive airway pressure, SVT = surfactant
The sodium level of the whole group of VLBWI with sodium abnormalities is shown in Fig 2. There was a decline in sodium levels over the first 5 days of life.
Hypernatraemia vs. Hyponatraemia
Table 2 compares the outcome between VLBWI with hypernatraemia and hyponatraemia. Although there were very few infants with hyponatraemia in the study, mortality was significantly higher in those VLBWI with hyponatraemia as compared to hypernatraemia (p = 0.019). Significantly more VLBWI with hyponatraemia had NEC than those with hypernatraemia (p < 0.001).
Comparison between VLBWI with hypernatraemia and hyponatraemia.
| Hyponatraemia | Hypernatraemia | p-Value | |
|---|---|---|---|
| Died | 5/6 (83.3%) | 25/79 (31.6%) | 0.019 |
| Necrotising enterocolitis | 4/6 (66.7%) | 1/79 (1.3%) | <0.001 |
| Antenatal steroids | 4/5 (80%) | 27/79 (34.2%) | 0.06 |
| Nasal continuous positive airways pressure | 2/6 (33.3%) | 60/79 (75.9) | 0.043 |
Mortality within the hypernatraemia group
Survivors within the hypernatraemia group were significantly greater than non-survivors 1145.0 g (SD 204.1) vs. 855.6 g (SD 189.7) (p < 0.001) and of more advanced gestational age 29.6 weeks (SD 2.4) vs. 26.8 weeks (SD 2.1) (p < 0.001). Other associations with mortality within the group of hypernatraemic VLBWI are shown in Table 3. Death was more likely in those VLBWI with concomitant hyperglycaemia, metabolic acidosis, NEC and birth by vaginal delivery. The use of NCPAP was associated with a reduced risk of death.
Significant associations with mortality within the hypernatraemic group of VLBWI.
| Variable | Survived | Died | p-Value |
|---|---|---|---|
| Hyperglycaemia | 7/54 (13%) | 13/25 (52%) | 0.001 |
| NEC | 2/54 (3.7%) | 6/25 (24%) | 0.011 |
| Metabolic acidosis | 0/54 (0) | 3/25 (12.0) | 0.029 |
| Nasal CPAP | 45/54 (83.3) | 15/25 (60) | 0.024 |
| NVD | 21/54 (38.9) | 17/23 (73.9) | 0.005 |
Fig 3 illustrates the difference in sodium levels between survivors and non-survivors in hypernatraemic VLBWI. Overall non-survivors had higher sodium levels than those who survived.
All sodium levels were significantly higher (p < 0.005), apart from the levels on day 4 which were statistically not different.
Discussion
The present study found that 28% of VLBWI had sodium abnormalities in the first week of life. Hypernatraemia was the most common sodium abnormality, present in 95% of the VLBWI with sodium abnormalities. Hypernatraemia was present in 26% of all VLBWI which is in keeping with previous reports.(20) Very few VLBWI in the present study were found to have hyponatraemia in the first week of life. This is in contrast to other reports where hyponatraemia in the first week of life was found to occur in 25%–65% of VLBWI in the first week of life and was more frequent in those infants of lower gestational age.(8,11)
The reason for the low frequency of hyponatraemia in the current study is unclear, but may reflect the practices in the neonatal unit during the study period. Most VLBWI were nursed under overhead radiant warmers during the study period and it is possible that methods to prevent insensible water loss, such as covering the infants with plastic sheeting, were not rigorously applied, resulting in frequent hypernatraemic dehydration and low occurrence of hyponatraemia. Renal function was not recorded in the study, but it is possible that VLBWI with renal dysfunction and reduced urine output, developed dilutional hyponatraemia as the standard maintenance fluid used in the unit had a very low sodium content.
In the current study, death in VLBWI with hypernatraemia was associated with lower gestational age and birthweight as well as concomitant hyperglycaemia, NEC and metabolic acidosis.
Serum sodium levels reflect water and sodium balance which changed over time after birth. Water loss through the skin in VLBWI, so-called ‘insensible water loss’ is high, resulting in hypernatraemia.(21) This insensible water loss is high post-delivery in VLBWI and decreases with increased post-natal age.(22) In the present study, it was impossible to evaluate fluid intake or insensible water loss as a possible cause of hypernatraemia as such data was not collected in this study. The majority of VLBWI in the current study treated with NCPAP had hypernatraemia. In the study unit, VLBWI who receive nasal CPAP are nursed on open resuscitation cribs with overhead radiant warmers which may have increased insensible water loss resulting in hypernatraemia. Hyperglycaemia and metabolic acidosis were significantly associated with mortality in hypernatraemic infants in the present study. Both hyperglycaemia and acidosis could be markers of hypernatraemic dehydration. In VLBWI with hypernatraemia, the presence of hyperglycaemia increased the risk of early mortality and IVH.(23) Phototherapy has been associated with increased insensible water loss, but was not found to be the case in this study.
There is a strong association between insensible water loss and birthweight.(24) The present study reflected that there was no significant association between sodium level and birthweight or gestational age. A study conducted in extremely preterm infants also found no association between birthweight and hypernatraemia.(5) Those VLBWI who died in the hypernatraemia group in the present study, however, were smaller and of lower gestational age than survivors.
This study has shown that low sodium levels were associated with an increased risk of mortality. This should be interpreted with caution due to the low number of subjects, many of whom had NEC, which may be an additional confounder. Serum sodium levels were higher in VLBWI who died within the hypernatraemic group.
Several studies have shown an association between hypernatraemia and IVH.(2,7) However, in a study by Hye Jin Lee et al.,(25) there was no significant association between IVH and hypernatraemia. There was no association between IVH and sodium level in the present study.
In the present study, there was a trend towards VLBWI who had received antenatal steroids to develop hyponatraemia as opposed to hypernatraemia. Administration of antenatal steroids and NEC was more frequent in VLBWI with hyponatraemia. Antenatal steroids have been reported to decrease the insensible water loss and decrease the incidence of hypernatraemia (26), PDA (12) and NEC (15), which are all associated with hypernatraemia in VLBWI.
Study limitations
This was a retrospective study and there was a small number of hyponatraemia patients reported. There was no information on fluid intake, type of incubator, cardiovascular status or the use of plastic shields which all influence sodium balance. It could not be determined whether sodium imbalance alone was the cause of death in the present study.
Conclusion
In conclusion, the present study showed a high incidence of sodium imbalance in VLBWI in the first week of life. Hypernatraemia was far more common than hyponatraemia. Mortality within the hypernatraemic infants was significantly associated with higher sodium level, lower birthweight and lower gestational age. Measures to prevent hypernatraemia in VLBWI should be implemented in the first week of life, particularly in very small premature infants.