Maternal intrapartum fluids and neonatal weight loss in the breastfed infant

Reduction in weight is a physiological process (Kellams et al, 2017), as an infant adjusts from dependence on maternal circulation for continuous nutrition to managing their own intermittent nutritional intake. Weight nadir usually occurs between days 2 and 4 after birth, recovering to birth weight by the end of the second week (National Institute for Health and Care Excellence, 2017). Acceptable parameters of loss remain unclear in the literature (Rich, 2022) but 10% loss from birth weight is generally cited as a threshold for intervention (National Institute for Health and Care Excellence, 2017). In the context of significant weight loss in a breastfed infant, efficacy of milk transfer becomes a central concern for clinicians (Jayaraj et al, 2020).

Administration of intravenous fluids during the intrapartum period is common practice when anaesthesia is required or complications occur. Intrauterine fluid resuscitation is routinely undertaken to support fetal wellbeing when compromise is suspected (Mahfouz et al, 2022). Transplacental passage of fluids administered to a mother intravenously in the intrapartum period appears to have an association with neonatal birth weight and subsequent loss through diuresis in the early days of life (Mulder and Gardner, 2015; Deng and McLaren, 2018).

Significant weight loss in a breastfed infant during the first weeks of life can be a serious health concern, and is a key factor in clinical decision making regarding the need for supplementation to avoid associated complications, such as hypernatraemia. Supplementation is a significant risk factor for early cessation of breastfeeding (Deng and McLaren, 2018), and this is an important public health issue globally (Victora et al, 2016; World Health Organization and United Nations Children's Fund, 2021; Wang et al, 2022; Pérez-Escamilla et al, 2023).

Recognising this phenomenon as a differential cause of weight loss may support identification of genuinely unwell infants, allowing targeted implementation of practices that support the continuation of exclusive breastfeeding when illness can be excluded. This review's objective was to identify whether maternal fluid balance/load is associated with neonatal weight loss after birth.

Methods

Searches of key databases (CINAHL, MEDLINE, EMBASE, EMCARE) were conducted using a search strategy developed in collaboration with the local NHS library service (Table 1). Known researchers in this area were contacted regarding unpublished/ongoing work of relevance, and reference lists of searched papers were examined backwards and forwards for relevant papers. Conference proceedings were also reviewed for relevant unpublished work. No date range limitation was applied to maximise retrieval of relevant papers, and all types of study were eligible for inclusion; however, papers not published in English were excluded because of the lack of translation facility.

Search results

Population-Exposure-Outcome searches, all combined with AND, gave a total of 240 records (MedLine: 39, CINAHL: 30, EmBase: 107, EMCare: 64). After being entered into Covidence for de-duplication, 154 records were retained. Two authors screened the papers by title/abstract, and all study methodologies meeting the inclusion criteria were considered for the study (Table 2), with 14 papers being retained. The main reasons for exclusion after screening related to population criteria; for instance, neonatology studies of premature or low birth weight babies. Other exclusions were made on the basis of intervention criteria, where fluid balance or weight loss was studied during the later neonatal period in relation to nutritional intake. The PRISMA flow diagram shown in Figure 1 outlines the full screening process.

Data extraction

Data were extracted from these papers using a structured form based on the Cochrane handbook for systematic reviews, which is available from the authors on request. All studies were assessed using the Cochrane risk of bias tool, and the results are illustrated in Figure 2. This process was independently reviewed by a supervisor of the lead author.

It was expected that there would be relatively few studies in this area and approaches to data collection, intervention definition and outcome selection would be heterogenous in nature, because of differences in clinical criteria across settings. It was therefore not anticipated that meta-analysis would be possible, and a descriptive review was conducted.

Quality of the review

Every effort was made to minimise bias during the present review process. Support was obtained from NHS library services to ensure access to a wide range of sources and standardised approaches to data extraction and reporting were used. Screening and data extraction were completed by two team members to minimise bias, and conclusions were discussed throughout the review team before publication.

Despite efforts to ensure comprehensive searching of the literature, relevant papers were noted after the review was completed. A prospective observational study conducted in Jerusalem (Rizik et al, 2018) found that intravenous fluid contributed to the multifactorial process of newborn weight loss among 375 women–baby pairs in this setting; and extensive work was also carried out by Mulder in this area (Mulder et al, 2010; Mulder and Gardner, 2015), which suggested an influential relationship between maternal fluid status and newborn weight loss.

Results and discussion

Description of studies

Population

All included studies examined outcomes for singleton infants, and all referred to ‘healthy’ infants. However, ‘health’ was defined in a variety of ways, from the absence of undefined significant perinatal morbidity (John et al, 2017) to ‘mother and baby discharged together and able to breastfeed freely’ (Noel-Weiss, 2011; Giudicelli et al, 2020).

All studies excluded infants born before 37 weeks of pregnancy, and Thulier (2013) excluded any born before 38 full weeks' gestation. Most studies reported on the basic demographic characteristics of their sample. The majority of participants were white, English-speaking, and educated to basic secondary standard. All studies took place in western healthcare settings. The total number of participants included was 1665.

Intervention

The focus of all studies was administration of intravenous fluids given during the intrapartum period. However, each study defined this intervention slightly differently. For instance, some studies included fluids administered from admission in labour/pre-caesarean until birth, while Chantry et al (2011) attempted to create a sub-group to examine the effect of fluids given during the 2 hours immediately prior to birth. This was a change to the original protocol, implemented after recruitment had commenced, when electronic records made such data collection possible, and consequently only included 63.8% of the study population.

Comparison

Comparison groups differed across the studies included in this review. Groups were categorised by total intravenous fluid administered in six studies, and according to ml/hour in the others. Differing thresholds were used for comparison, reflecting heterogeneity in exposure. Chantry et al (2011) compared intravenous loads of <100ml/h, with 100–200ml/h and >200ml/h, while others dichotomised using thresholds of 1200ml total (Noel-Weiss, 2011), 1500ml (Giudicelli et al, 2020), and even 3000ml (Thulier, 2013). In Watson's (2010) randomised controlled trial, a ‘conservative’ regime of intravenous fluid administration was compared against usual care. These both considered the effects of bolus loading and ongoing ml/hour infusions. Minimal reference was made to oral intake of fluids.

Outcomes

All studies included outcomes related to infant weight loss in either percentage terms or absolute loss in grams. The definition of ‘excessive’ weight loss was variable across the studies, and ranged from 7% loss (Watson, 2010) to 10% (Chantry et al, 2011). These studies, along with Thulier (2013) compared the prevalence of ‘excessive’ weight loss according to intravenous administration, while others compared mean weight loss (Noel-Weiss, 2011; Buonafede et al, 2019; Giudicelli et al, 2020). Lamp and Macke (2010) used regression modelling to examine the relationship between weight loss (g) to other variables of interest, including maternal fluids.

Weight loss was examined at various times, with all studies including measurements at specified time-points during the first 4 days of life. Thulier's (2013) study recorded weights on certain days, and did not take into account the age in hours of the infant, a limitation identified by the author. Lamp and Macke (2010) recorded weight at 24 and 48 hours exactly. John et al (2017) collected weight data 12-hourly until 72 hours of age, and then again on ‘day 10’ of life.

No two studies in the review collected the same outcome data at the same time, using the same comparator or intervention. Of the eight studies presenting original data from 1665 participants, four reported an association between maternal intravenous fluid intake and neonatal weight loss, and four did not. Table 3 shows a summary of the included study characteristics.

Quality of the evidence

The majority of studies were observational, which resulted in bias on a number of levels. Poor reporting prevented a full assessment of methodological rigour; however, the use of maternally-reported outcome data was noted to introduce measurement bias and attrition bias was an issue for at least one study (Chantry, 2011). The narrative papers reviewed did not take systematic approaches to literature review. The one randomised controlled trial in this review was conducted using robust methods with a low level of bias. It did not observe an effect of restricted intravenous administration on neonatal weight loss. However, the authors posed the question of whether investigation of total maternal volumes in excess of 2500ml may be of value in future research.

In a systematic review of physiological newborn weight loss research in 2008, Noel-Weiss et al (2008) found that the methods used to report weight loss were inconsistent, and that this was a barrier to clear investigation of the relationship between weight loss and morbidity. The present review notes that this problem persists today, precluding observation of the nature of a potential relationship between maternal fluid status and neonatal weight loss.

All the studies reflected predominantly white, western populations and healthcare settings, restricting the extent to which their findings can be generalised to more diverse settings.

Implications for practice

Despite conflicting conclusions among studies in this review, all authors suggested that attention should be paid to maternal fluid intake when assessing significant weight loss in a breastfed infant. While it remains impossible to apply specific thresholds for concern because of limitations in the collection of research evidence, it is generally recognised that a relationship of some nature exists. It is possible that mothers and babies could benefit from clinicians considering this aspect of clinical history in postnatal assessments.

Implications for research

Further research is required in this area to clarify the relationship between maternal fluid intake and neonatal weight loss in a way that is clinically useful.

Measuring total intravenous fluid volume received during the intrapartum period would allow comparison to existing studies, and may help to confirm the observations of other authors around a potential ‘threshold’ above which maternal fluid load becomes relevant to neonatal weight loss. Since the definition of ‘excessive’ or ‘significant’ weight loss in the newborn remains a contentious issue, it may be more helpful to examine neonatal weight loss as a continuous measure in grams or percentage terms in preference to varying categories of excessive weight loss, as this limits the extent to which comparisons can be made.

Conclusions

This review demonstrated that attention must be paid to selection of outcome and method of measurement to allow for comparison with existing evidence that can lead to a meaningful body of data to inform clinical practice. Given the importance and urgency of the public health issues associated with the premature cessation of breastfeeding, it is imperative this area of study remains in the spotlight.

Key points

• Transplacental passage of fluids intravenously administered to a mother in the intrapartum period may have an association with neonatal birthweight and subsequent loss through diuresis in the early days of life.
• Recognition of such non-pathological causes of weight loss may support the identification of genuinely unwell infants, thereby targeting practices that better support continuation of exclusive breastfeeding when illness can be excluded.
• The definition of ‘excessive’ or ‘significant’ weight loss remains a contentious issue, and methods used to report weight loss remain inconsistent, preventing clear comparison of data.
• Further research is required in this area to clarify the relationship between maternal fluid intake and neonatal weight loss in a way that is clinically useful.
• It is possible that a ‘threshold’ above which maternal fluid load becomes relevant to neonatal weight loss can be identified for use in clinical practice.

CPD reflective questions

• Does your clinical area have guidance on intravenous fluid volumes during the intrapartum period?
• What is a ‘significant’ weight loss in your clinical area? How does this fit with your opinion/observations in practice?
• What factors might be taken into account when considering significant weight loss in an otherwise well-breastfed infant?
• How might this paper influence your assessment of breastfed infants in future?