Systematic review of lipid emulsions for intravenous nutrition in term and late preterm infants.

Review question: which lipid (fat) emulsions (LE) have the best outcomes in term and late preterm infants (greater than 34 weeks' gestation (time between becoming pregnant and giving birth)) with or without pre-existing liver disease and surgical conditions?

Background: infants who need nutrition through an intravenous line (tube into a vein) have been conventionally given pure soybean oil-based lipid emulsions. However, high polyunsaturated fatty acid (PUFA) content and phytosterols in pure soybean oil-based LEs may be harmful and may contribute to parenteral nutrition-associated liver disease (PNALD) and other side effects. The newer LEs from other lipid sources including fish oil may potentially improve clinical outcomes in term and late preterm infants by decreasing PUFA content and providing specific benefits.

Study characteristics: the review authors searched the medical literature and identified nine eligible studies (including 273 preterm infants). Evidence is up to date as of 18 June 2018.

Key findings: there is a paucity of research in this field with a limited number of studies, many of which were terminated early. There is very low quality and limited evidence to suggest that fish oil-based LE may improve some liver disease-related outcomes in infants with pre-existing liver disease. However, this evidence was based on very limited number of participants from two small studies, one of which was terminated early, so that no certain conclusions could be drawn. There was no evidence of benefit in other clinically important outcomes, for example, sepsis, rate of growth or death with any LE including fish oil-LE over another LE in term and late preterm infants with surgical conditions or liver disease. We found no studies in infants without surgical conditions or pre-existing liver disease.

Conclusions: we found that in the population of term and late preterm infants (greater than 34 weeks' gestation) with pre-existing liver disease or surgical conditions, there is currently insufficient evidence from well-designed studies to determine beneficial effects of any LE including the fish oil-containing LEs over another LE for prevention or resolution of liver disease or other clinically important outcomes.

Authors' conclusions: 

Based on the current review, there is insufficient data from randomised studies to determine with any certainty, the potential benefit of any LE including fish oil-containing LEs over another LE, for prevention or resolution of PNALD/cholestasis or any other outcomes in term and late preterm infants with underlying surgical conditions or cholestasis. There were no studies in infants without surgical conditions or cholestasis.

Further research is required to establish role of fish oil or lipids from other sources in LEs to improve PNALD/cholestasis, and other clinical outcomes in parenterally fed term and late preterm infants.

Read the full abstract...

Lipid emulsions (LE) form a vital component of infant nutrition for critically ill, late preterm or term infants, particularly for those with gastrointestinal failure. Conventionally used soybean oil-based LE (S-LE) have high polyunsaturated fatty acid (PUFA) content and phytosterols, which may contribute to adverse effects including parenteral nutrition-associated liver disease (PNALD).


To compare the safety and efficacy of all LE for parenteral nutrition (PN) in term and late preterm infants (between 34 weeks' gestation and 36 weeks' and six days' gestation) with or without surgical conditions or PNALD within first six months of life, using all possible direct comparisons.

Search strategy: 

We used the standard search strategy of Cochrane Neonatal to search the Cochrane Central Register of Controlled Trials (CENTRAL 2018, Issue 5), MEDLINE (1946 to 18 June 2018), Embase (1974 to 18 June 2018), CINAHL (1982 to 18 June 2018), MIDRIS (1971 to 31 May 2018), conference proceedings, trial registries ( and the WHO's Trials Registry), and the reference lists of retrieved articles for randomised controlled trials and quasi-randomised trials.

Selection criteria: 

Randomised or quasi-randomised controlled studies in term and late preterm infants, with or without surgical conditions or PNALD.

Data collection and analysis: 

Data collection and analysis conformed to the methods of Cochrane Neonatal. We used the GRADE approach to assess the quality of evidence for important outcomes in addition to reporting the conventional statistical significance of results.

Main results: 

The review included nine randomised studies (n = 273). LE were classified in three broad groups: 1. all fish oil-containing LE including pure fish oil (F-LE) and multisource LE (e.g. medium-chain triglycerides (MCT)-olive-fish-soybean oil-LE (MOFS-LE), MCT-fish-soy oil-LE (MFS-LE) and olive-fish-soy-LE (OFS-LE)); 2. conventional pure S-LE; 3. alternative-LE (e.g. MCT-soy-LE (MS-LE), olive-soy-LE (OS-LE) and borage oil-based LE).

We considered four broad comparisons: 1. all fish oil LE versus non-fish oil LE (6 studies; n = 182); 2. fish oil LE versus another fish oil LE (0 studies); 3. alternative-LE versus S-LE (3 studies; n = 91); 4. alternative-LE versus another alternative-LE (0 studies) in term and late preterm infants (0 studies), term and late preterm infants with surgical conditions (7 studies; n = 233) and term and late preterm infants with PNALD/cholestasis (2 studies; n = 40).

PNALD/cholestasis was defined as conjugated bilirubin (Cbil) 2 mg/dL or greater and resolution of PNALD/cholestasis as Cbil less than 2 mg/dL. We put no restriction on timing of PNALD detection. There was heterogeneity in definitions and time points for detecting PNALD in the included studies.

We found one study each in surgical infants and in infants with cholestasis, showing no evidence of difference in incidence or resolution of PNALD/cholestasis (Cbil cut-off: 2 mg/dL) with use of fish oil-containing LE compared to S-LE.

We considered an outcome allowing for any definition of PNALD (different Cbil cut-off levels). In infants with surgical conditions and no pre-existing PNALD, meta-analysis showed no difference in the incidence of PNALD/cholestasis (any definition) with use of fish oil-containing LE compared to S-LE (typical risk ratio (RR) 1.20, 95% confidence interval (CI) 0.38 to 3.76; typical risk difference (RD) 0.03, 95% CI -0.14 to 0.20; 2 studies; n = 68; low-quality evidence). In infants with PNALD/cholestasis (any definition), use of fish oil-LEs was associated with significantly less cholestasis compared to the S-LE group (typical risk ratio (RR) 0.54, 95% confidence interval (CI) 0.32 to 0.91; typical risk difference (RD) –0.39, 95% CI –0.65 to –0.12; number needed to treat for additional beneficial outcome (NNTB) 3, 95% CI 2 to 9; 2 studies; n = 40; very low-quality evidence). This outcome had very low number of participants from two small studies with differences in study methodology and early termination in one study, which increased uncertainty about the effect estimates.

One study in infants with cholestasis reported significantly better weight gain with a pure fish oil LE compared to a 10% S-LE (45 g/week, 95% CI 15.0 to 75.0; n = 16; very low-quality evidence). There were no significant differences in growth parameters in studies with surgical populations.

For the secondary outcomes, in infants with cholestasis, one study (n = 24) reported significantly lower conjugated bilirubin levels but higher gamma glutamyl transferase levels with MOFS-LE (SMOFlipid) versus S-LE (Intralipid) and another study (n = 16), which was terminated early, reported significantly higher rates of rise in alanine aminotransferase (ALT) and conjugated bilirubin levels in the S-LE group compared to pure F-LE (Omegaven).

In surgical infants, two studies each reported on hypertriglyceridaemia and Cbil levels with one study in each outcome showing significant benefit with use of a F-LE and the other study showing no difference between the groups. Meta-analysis was not performed for either of these outcomes as there were only two studies showing conflicting results with high heterogeneity between the studies.

There was no evidence of differences in death, sepsis, alkaline phosphatase and ALT levels in infants with surgical conditions or cholestasis (very low-quality evidence).

One study reported neurodevelopmental outcomes at six and 24 months in infants with surgical conditions (n = 11) with no evidence of difference with use of pure F-LE versus S-LE. Another study in infants with cholestasis (n = 16) reported no difference in head growth velocity between pure F-LE versus S-LE.

GRADE quality of evidence ranged from low to very low as the included studies were small single-centre studies. Three of the six studies that contributed data to the review were terminated early for various reasons.