Neonatal parenteral nutrition: An expert narrative review of international guidelines and contemporary evidence

When to start neonatal parenteral nutrition

NICE provides the most explicit operational advice on when neonatal PN should begin. For babies who meet criteria for PN, it recommends starting as soon as possible, no later than 8 h, and defines feed-interruption thresholds that should prompt PN in both preterm and term infants. ² ASPEN supports similarly prompt initiation after birth once appropriate vascular access is available, although its document is directed specifically at preterm infants. ¹ The European guidance is also aligned with early nutritional support, but it tends to frame the issue in physiologic rather than workflow terms.^3–12 Taken together, the three approaches support an early-start strategy whenever enteral adequacy is unlikely. These operational differences are summarized in Table 1.

The area that remains genuinely controversial is the critically ill term neonate. NICE takes a distinctly neonatal position and provides nutritional ranges for term infants receiving PN, including those who are critically ill or post-operative, while advising clinicians to stay at the lower end of the energy range when appropriate. ² The European amino acid guidance is more cautious and allows consideration of withholding PN, particularly amino acids, during the acute phase of critical illness. ⁵ That caution is supported by the neonatal subgroup analysis of the PEPaNIC trial. In this preplanned subgroup of 209 term neonates enrolled before 4 weeks of age, late PN, started on day 8 rather than on admission, was associated with fewer new infections and earlier discharge from the pediatric intensive care unit, although hypoglycemia was more frequent. ¹⁴ The practical implication is that early PN remains standard for preterm infants and many surgical or feed-intolerant neonates, but the acutely critically ill term infant still requires individualized judgment rather than automatic protocolized escalation.

Energy and glucose

Energy prescription in neonatal PN is constrained by fluid allowance, substrate tolerance, and the infant’s rapidly changing metabolic state after birth. NICE recommends that when PN is started within the first 4 postnatal days, total parenteral energy should begin at 40–60 kcal/kg/day and then increase over roughly 4 days to 75–120 kcal/kg/day; if PN starts after day 4, the maintenance range is 75–120 kcal/kg/day from the outset. ² These ranges are practical because they are easy to translate into standardized formulations and everyday NICU prescribing. The main numerical ranges across guidelines are compared in Table 2. The European carbohydrate guidance approaches the same problem from a more physiologic angle by expressing glucose supply in mg/kg/min. For preterm newborns, it recommends starting at 4–8 mg/kg/min on day 1 and advancing over 2–3 days to a target of 8–10 mg/kg/min, with an upper maximum of 12 mg/kg/min. For term newborns, the starting rate is lower at 2.5–5 mg/kg/min, increasing to a target of 5–10 mg/kg/min. ⁷ This framing is especially helpful when bedside decisions are being driven by glycemic tolerance rather than by a pre-composed PN bag.

In practice, NICE and ESPGHAN are more complementary than conflicting. NICE is easier to operationalize in unit protocols and standardized bags, whereas the European guidance provides the metabolic guardrails that matter when infants do not follow an expected course. ELBW infants deserve particular caution because glucose intolerance, insulin exposure, osmotic diuresis, fluid restriction, and concurrent sepsis or steroid exposure may rapidly change glucose tolerance. The shared clinical principle is to avoid both underfeeding and carbohydrate overload, and to adjust glucose delivery to tolerance rather than to formula alone.

Amino acids

Amino acid provision is the most studied and most debated component of neonatal PN. All three major guidance systems now endorse early amino acid delivery, but they differ in how aggressively they escalate it. NICE recommends that preterm infants who start PN within the first 4 days receive 1.5–2 g/kg/day initially, increasing over about 4 days to 3–4 g/kg/day; for term infants, the suggested start is 1–2 g/kg/day with progression to 2.5–3 g/kg/day. ² ASPEN supports early amino acid initiation in preterm infants but advises avoiding an initial dose above 3 g/kg/day and avoiding routine doses above 3.5 g/kg/day. ¹ The European amino acid guideline recommends starting at 1.5–2.5 g/kg/day on day 1 and advancing to a usual target of 2.5–3.5 g/kg/day. ⁵ These ranges are summarized in Table 2.

The evidence behind that consensus matters because neonatal nutrition has moved away from the older idea that if some early amino acid is beneficial, then substantially more must be better. The updated Cochrane review comparing early versus later amino acid administration found better nitrogen balance with earlier administration, but did not show convincing improvement in mortality, growth, or neurodevelopmental outcomes. ¹⁶ Likewise, randomized data comparing higher with more moderate amino acid targets have not consistently shown superior long-term benefit, and the NEJM follow-up of early amino acid strategies in extremely preterm infants did not support a broad outcome advantage from more aggressive exposure.^17,18 The key studies shaping this interpretation are summarized in Table 4.

OPEN IN VIEWER

Table 4.

Selected trials, reviews, and evidence level informing current neonatal PN practice.

Study	Design/evidence type	Key finding	Practice implication
Stephens et al., 2009 13	Observational cohort of ELBW infants	Higher first-week protein and energy intake was associated with better 18-month developmental scores.	Supports attention to cumulative early nutrient deficits, while acknowledging observational design.
van Puffelen et al., 2018 14	PEPaNIC term-neonate subgroup analysis	Late PN in critically ill term neonates was associated with fewer new infections and earlier PICU discharge, but more hypoglycemia.	Term critical illness remains a setting where timing of PN should be individualized.
Trivedi et al., 2024 16	Cochrane review of early vs late amino acids	Early amino acids improved nitrogen balance, but evidence for improved mortality, growth, or neurodevelopment remained insufficient.	Supports early amino acids, but not excessive certainty that earlier always improves hard outcomes.
Burattini et al., 2013 17	Randomized trial, 2.5 vs 4 g/kg/day amino acids	Higher amino acid dosing increased biochemical differences but did not improve growth or neurodevelopment.	Routine escalation to 4 g/kg/day is not justified by outcome data.
Bloomfield et al., 2022 18	Randomized trial of additional early amino acids	Extra early amino acids did not improve survival without neurodisability at 2 years.	Reinforces current consensus ceilings and argues against ‘more is better’ amino acid strategies.
Vlaardingerbroek et al., 2012 19	Systematic review/meta-analysis of early lipids and lipid composition	Early lipids appeared safe and well tolerated but did not clearly improve growth or major morbidity.	Supports early lipid use, but not strong claims of broad superiority for any specific emulsion.
He et al., 2024 20	Meta-analysis of SMOFlipid RCTs	SMOFlipid reduced cholestasis risk without consistent changes in other major neonatal outcomes.	Composite emulsions may be most useful in prolonged PN or cholestasis-prone infants.
Zou et al., 2022 21	Meta-analysis in VLBW infants	Fish-oil-containing emulsions reduced PN-associated cholestasis, especially when PN lasted >14 days.	Strengthens selective rather than universal use of fish-oil-containing emulsions.
Mihatsch et al., 2023 25	Systematic review of standardized vs individualized PN	Standardized PN may increase nutrient delivery and improve growth, without clear effects on mortality or NEC.	Supports standardized PN as a systems intervention, while preserving individualization when needed.
Frost et al., 2023 and Gabriel et al., 202323,24	Review and national survey	No universal definition or consistent management approach exists for neonatal hypertriglyceridemia.	Triglyceride values should be interpreted in clinical context rather than as an automatic stop-lipid rule.
Cairns and Stalker, 2000 26	Cochrane review	No evidence of benefit from routine carnitine supplementation for weight gain, lipid utilization, or ketogenesis.	Routine prophylactic carnitine is not supported for default PN protocols.

A second lesson from the contemporary literature is that amino acid dosing cannot be interpreted in isolation. Phosphate, potassium, magnesium, and adequate non-protein energy all matter. Reviews of protein-related controversies in very preterm infants emphasize that increased amino acid exposure can precipitate refeeding-like hypophosphatemia and related biochemical disturbances when the anabolic drive increases without concurrent mineral support.^5,8,12 This is particularly important in ELBW infants, who are most vulnerable to cumulative protein deficits but also to phosphate depletion and glucose intolerance.

Clinically, amino acid strategies should be judged against meaningful neonatal outcomes, not only blood urea nitrogen, acid-base status, or short-term nitrogen balance. Avoiding cumulative protein-energy deficit, preserving head growth, reducing postnatal growth failure, and improving long-term neurodevelopment remain the outcomes that matter most. Observational data link better early protein and energy intake with later developmental scores in ELBW infants, but randomized evidence has not yet defined a universally superior high-dose threshold.^13,16–18 A cautious synthesis is therefore to start amino acids early, avoid unnecessary under-provision, but not to escalate beyond contemporary targets unless the rest of the anabolic package and biochemical monitoring are secure.

Lipids and lipid emulsion choice

Early lipid administration is now standard practice in neonatal PN. Lipids are energy-dense, iso-osmolar, and essential for preventing essential fatty acid deficiency. NICE recommends starting lipids at 1–2 g/kg/day and increasing by 0.5–1 g/kg/day to a maintenance range of 3–4 g/kg/day. ² ESPGHAN treats intravenous lipid emulsion as an integral part of PN and recommends that parenteral lipid intake generally should not exceed 4 g/kg/day in preterm and term infants. ⁶ Table 2 provides a numerical comparison of lipid dosing recommendations.

The more difficult issue is no longer whether lipids should be used, but which lipid emulsion to prefer. The older meta-analysis by Vlaardingerbroek and colleagues suggested that introducing lipids within the first 2 days of life is safe and well tolerated, but did not demonstrate clear benefits for growth or major morbidity. ¹⁹ It also found only limited evidence that newer emulsions outperform pure soybean-oil emulsions across broad neonatal outcomes.

More recent evidence suggests a narrower, but clinically meaningful, advantage for some composite emulsions. A 2024 meta-analysis of randomized trials reported that SMOFlipid reduced the risk of cholestasis compared with other emulsions, but did not significantly affect mortality, sepsis, necrotizing enterocolitis, bronchopulmonary dysplasia, patent ductus arteriosus, retinopathy of prematurity, or length of stay. ²⁰ A 2022 meta-analysis of fish-oil-containing emulsions similarly found a preventive effect against PN-associated cholestasis in very low birth weight infants, especially when PN lasted longer. ²¹ These findings support considering mixed-oil or fish-oil-containing emulsions in infants at an increased risk of prolonged PN or liver complications.

At the same time, caution remains necessary. Consistent superiority across major neonatal outcomes has not been demonstrated, and lipid composition may influence arachidonic acid, docosahexaenoic acid, and eicosapentaenoic acid exposure in ways that are not yet fully understood.^6,22,27 Essential fatty acid content also differs across products; therefore, the lipid dose required to prevent essential fatty acid deficiency may differ by emulsion. ESPGHAN recommends a minimum linoleic acid intake of 0.25 g/kg/day in preterm infants and 0.1 g/kg/day in term infants, but translating this into bedside dosing is not straightforward because products vary in linoleic acid content.^6,22

A further operational issue is medication compatibility. There is substantially more accumulated compatibility experience with standard soybean-oil lipid emulsions than with mixed-oil or fish-oil-containing products. Compatibility information for soybean-oil emulsions should not automatically be applied to other lipid emulsions unless supportive data are available; pharmacists and product-specific compatibility information should guide coadministration decisions.^15,27 A clearer practical recommendation is therefore: use early lipids routinely, do not assume that one emulsion improves all neonatal outcomes, consider composite or fish-oil-containing emulsions when prolonged PN or cholestasis risk is high, and interpret triglycerides in clinical context rather than as a single automatic trigger for lipid cessation.

Fluids, electrolytes, calcium, phosphate, and magnesium

Electrolyte and mineral prescribing is where physiology often matters most. Neonatal PN is delivered during a period of rapid extracellular fluid contraction, renal adaptation, and abrupt interruption of placental mineral transfer. NICE provides practical bedside ranges, recommending calcium 0.8–1.0 mmol/kg/day initially and 1.5–2 mmol/kg/day as maintenance, together with phosphate 1 mmol/kg/day initially and 2 mmol/kg/day as maintenance. ² These values provide clinicians with a workable framework for early prescribing and are summarized in Table 3.

The European fluid and electrolyte guideline adds an important phase-based perspective. It recommends gradual increases in fluid intake after birth and notes that sodium, chloride, and potassium are often not required during the initial contraction phase associated with early postnatal weight loss. ⁸ In ELBW and VLBW infants, sodium and potassium may be introduced later depending on clinical status, weight trend, urine output, and laboratory data. This approach is especially useful when early fluid shifts are pronounced and when fixed maintenance targets could be misleading.

Magnesium attracts less debate, but it should not be treated as an afterthought. NICE recommends providing magnesium ideally from the outset, or as soon as possible after PN is started. ² In current practice, that point has become even more relevant because phosphate deficiency and refeeding-like syndromes rarely occur alone; they are commonly accompanied by hypokalemia and hypomagnesemia.^8,12 The key lesson from both guidelines and recent clinical practice is that early biochemical surveillance has to accompany any strategy that promotes anabolism, especially in ELBW, growth-restricted, or severely ill infants.

Vitamins, trace elements, and iron

Micronutrient guidance is one area where the major documents are closely aligned, even though the evidence base is less robust than it is for macronutrients. NICE recommends daily intravenous administration of water-soluble and fat-soluble vitamins, ideally from the start of PN or as soon as possible after initiation, and advises delivering vitamins within the lipid emulsion to improve stability. ² It also recommends daily trace elements and magnesium when PN is expected to continue for more than a few days. ESPGHAN similarly recommends parenteral vitamins and trace elements for infants receiving PN.^9,10 Practical ranges and monitoring considerations are summarized in Table 3.

The European trace element paper is more explicit about dosing. For preterm infants, it recommends zinc 400–500 microgram/kg/day, copper 40 microgram/kg/day, and iodine 1–10 microgram/kg/day, with closer monitoring during prolonged PN or in infants with high gastrointestinal losses. ⁹ Iron is handled differently from the other micronutrients. NICE advises against intravenous iron in babies younger than 28 days, whereas the European guidance discusses iron mainly in the context of longer-duration PN and special clinical circumstances.^2,9 In routine short-duration neonatal PN, early intravenous iron is therefore generally avoided.

The evidence gap is important. Many micronutrient recommendations are based on physiologic plausibility, deficiency prevention, expert consensus, and limited neonatal outcome data rather than large randomized trials. Therefore, prolonged PN, intestinal failure, high stoma losses, cholestasis, renal dysfunction, or unexplained poor growth should prompt individualized review of trace element and vitamin adequacy rather than reliance on a fixed default formulation.

Operational aspects

Monitoring and discontinuation

All major guideline systems treat monitoring as a part of PN itself rather than as an optional add-on. NICE is the most practical for bedside scheduling: it recommends glucose monitoring when PN begins and at every bag change; daily serum triglycerides while the lipid dose is being increased, and weekly once maintenance dosing is established; and daily phosphate monitoring during escalation, with less frequent review thereafter. ² ESPGHAN adds helpful threshold-based thinking, especially for dysglycemia, and reinforces the idea that monitoring intensity should reflect instability, prior abnormalities, and recent prescription changes.^7,8,12 A practical monitoring and discontinuation summary is provided in Table 5.

OPEN IN VIEWER

Table 5.

Practical monitoring and discontinuation summary for neonatal PN.

Domain	Early/escalation phase	Stable maintenance phase	When to individualize more closely
Glucose	Check at PN initiation, after bag changes, and after GIR changes.	Continue according to unit protocol and clinical stability.	ELGAN/ELBW, sepsis, steroids, insulin use, hypoglycemia risk, fluid restriction.
Triglycerides	Commonly checked during lipid escalation; NICE suggests daily while increasing lipids.	Often weekly once stable, but practices vary.	ELBW, SGA, sepsis, high lipid dose, cholestasis, suspected EFAD, sampling during infusion.
Phosphate, potassium, magnesium	Check frequently during amino acid and mineral escalation.	Reduce frequency once stable and receiving consistent intake.	ELBW, growth restriction, high amino acid advancement, refeeding-like risk, renal dysfunction.
Calcium and bone minerals	Review with phosphate during early mineral provision.	Monitor according to duration of PN and metabolic bone risk.	Very preterm infants, prolonged PN, diuretics/steroids, low phosphate, poor growth.
Liver function and bilirubin	Baseline and periodic review if PN persists.	Regular review during prolonged PN.	PN >2 weeks, intestinal failure, sepsis, cholestasis, trace element adjustment.
Trace elements/vitamins	Provide early if PN is expected to continue beyond a few days.	Routine broad levels are not required for every short PN course.	Prolonged PN, cholestasis, high losses, renal dysfunction, poor growth, suspected deficiency.
Discontinuation	Taper as enteral feeds advance; avoid abrupt nutrient gaps.	Stop when enteral intake reliably meets energy/protein needs and biochemistry is stable.	ELBW infants transitioning to fortification, recurrent feed interruptions, hypoglycemia risk.

Sources:^{1,2,7–12,22–24}.

A sensible neonatal monitoring bundle includes daily assessment of fluid balance and weight trend; close glucose surveillance during initiation and whenever dextrose concentration changes; frequent review of triglycerides, calcium, phosphate, magnesium, and potassium during early advancement; and continued monitoring of liver function and trace element status in infants who remain on PN for longer periods. ELBW, growth-restricted, septic, steroid-exposed, renal-impaired, and cholestatic infants require closer individualized review than stable larger preterm infants.

Stopping PN is much less contentious. As enteral nutrition advances, PN should be tapered progressively rather than withdrawn abruptly, especially in infants who still depend on PN for a meaningful proportion of energy or protein intake. The safest transition is usually staged, with biochemical stability and enteral tolerance improving together. During this transition, clinicians should avoid creating a protein-energy gap by stopping PN before enteral feeds and fortification can meet nutritional targets.

Current controversies and areas without full agreement

Several practical areas remain unsettled despite the availability of major guidelines. These controversies are summarized in Table 6. The first is triglyceride monitoring. NICE provides a monitoring schedule, but ASPEN does not provide an equivalent neonatal bedside schedule, and the biologic interpretation of triglyceride levels in very preterm infants remains uncertain.^1,2,23,24 Hypertriglyceridemia may reflect immature lipid clearance, illness severity, sepsis, growth restriction, sampling during infusion, or rarely essential fatty acid deficiency; therefore, an elevated triglyceride level should not automatically lead to complete lipid cessation. It should trigger clinical review of dose, infusion rate, sampling timing, sepsis, bilirubin risk, and the need to preserve essential fatty acid intake.^22–24

OPEN IN VIEWER

Table 6.

Current controversies and practical synthesis.

Controversy	Why disagreement remains	Practical synthesis
Triglyceride monitoring	No universal neonatal threshold; values may reflect immaturity, illness, sampling timing, or EFAD risk.	Monitor during escalation and in high-risk infants; interpret in context; avoid complete lipid cessation unless clearly indicated.
Hyperglycemia in ELGANs	Guidelines do not fully standardize how low GIR should be reduced before insulin.	Review GIR, sepsis/stress, drugs, and sampling; use insulin selectively with close glucose and potassium monitoring.
PN-associated cholestasis	Composite/fish-oil emulsions reduce cholestasis in some data but do not improve all outcomes.	Reduce PN duration, prevent sepsis, review trace elements, and consider composite/fish-oil lipids in prolonged PN or cholestasis.
Vitamin D, micronutrient, and EFA monitoring	Routine broad testing for all short PN courses is not evidence-based.	Test selectively in prolonged PN, intestinal failure, cholestasis, poor growth, or suspected deficiency.
Prophylactic carnitine	Low levels can occur during prolonged PN, but trials have not shown consistent clinical benefit.	Do not use routinely; consider only for prolonged PN, documented deficiency, or metabolic specialist indication.
Refeeding-like syndrome	Higher early amino acid delivery can unmask phosphate, potassium, and magnesium deficits.	Build PN protocols that link amino acid advancement to early phosphate, potassium, magnesium, and glucose surveillance.
Lipid emulsion composition	Newer emulsions may reduce cholestasis but do not show consistent superiority for growth, sepsis, NEC, BPD, ROP, or mortality.	Use early lipids routinely; select emulsion by cholestasis risk, PN duration, essential fatty acid needs, and compatibility data.

Sources:^{1,2,5–10,12,15,20–24,26,27}.

The second controversy is hyperglycemia in ELGANs. Guidelines agree that glucose should be titrated to tolerance, but they do not fully standardize how far to reduce glucose infusion rate before insulin is used.^1,2,7 In practice, the first steps are to confirm the glucose value, review GIR and concurrent medications, assess sepsis or stress physiology, and avoid excessive carbohydrate delivery. Insulin may be appropriate for persistent clinically significant hyperglycemia when adequate calories cannot otherwise be provided, but it should be used with careful monitoring because hypoglycemia and potassium shifts can be harmful. The third controversy is PN-associated cholestasis. Strategies include minimizing PN duration, advancing enteral feeds when safe, preventing catheter-related sepsis, cycling PN in selected longer-term patients, reviewing copper and manganese exposure in cholestasis, ensuring adequate zinc in infants with high losses, and considering composite or fish-oil-containing emulsions when PN is prolonged or cholestasis has developed.^6,9,12,20,21 The evidence supports selective use rather than a universal claim that newer lipid emulsions improve all neonatal outcomes.

The fourth controversy is micronutrient, vitamin D, and essential fatty acid monitoring. Routine frequent measurement of all micronutrients is not feasible or evidence-based for every short-term PN course. However, prolonged PN, intestinal failure, cholestasis, poor growth, unusual skin findings, recurrent infections, or restricted lipid dosing should prompt individualized testing, including consideration of fatty acid profile or triene:tetraene ratio when essential fatty acid deficiency is suspected.^9,10,22 The fifth controversy is prophylactic carnitine. Preterm infants may develop low carnitine levels during prolonged PN, but a Cochrane review found no evidence that routine supplementation of parenterally fed neonates improves weight gain, lipid utilization, or ketogenesis. ²⁶ Therefore, prophylactic carnitine should not be embedded as a routine default in neonatal PN protocols; it may be considered selectively in prolonged PN, documented deficiency, or specialist-directed metabolic contexts.