The study, led by researchers at NYU Langone Hospital – Long Island and Rainbow Babies and Children’s Hospital in Cleveland, followed 347 preterm infants born at or before 32 weeks of gestation with birth weights of 1500 grams or less. All of them required intravenous lipid emulsions, the fat component of parenteral nutrition, for at least seven consecutive days. Of these infants, 179 received the traditional 100% soybean oil emulsion, known as SO-ILE, while 168 received SMOF-ILE, a mixture of soybean oil, medium-chain triglycerides, olive oil, and fish oil that the US Food and Drug Administration approved for neonatal use only in 2022. The researchers compared changes in weight, length, and head circumference z-scores, standardized measures that track how far a baby sits from the average for their age, from birth to one month and from one month to 36 weeks postmenstrual age or discharge.

The biological stakes are high. Preterm infants miss the third-trimester period in which fat normally accumulates in utero, leaving them with dangerously low reserves of essential fatty acids. Deficiency can emerge within 72 hours of inadequate lipid intake. Linoleic acid and alpha-linolenic acid, the two essential fatty acids supplied by these emulsions, are converted in the body into arachidonic acid and, respectively, into eicosapentaenoic acid and docosahexaenoic acid, or DHA. All of these long-chain polyunsaturated fatty acids are indispensable for neurological, immune, and retinal development. The trouble is that soybean oil emulsions, despite supplying the raw precursors, contain little of the ready-made DHA and EPA that the developing brain and retina preferentially use.

That biochemical gap is precisely what motivated the development of mixed-oil emulsions. SMOF-ILE delivers substantially more omega-3 fatty acids, more antioxidant vitamin E, and fewer phytosterols, plant compounds that can impair immune function and interfere with bile acid secretion. Soybean oil emulsions at the recommended dose of 3 grams per kilogram per day have been linked to parenteral nutrition-associated cholestasis, a potentially serious liver injury, through mechanisms that include a high omega-6 load driving proinflammatory prostaglandins and leukotrienes. On paper, the newer blend looks like a clear upgrade. Yet clinical trials comparing the two have produced stubbornly inconsistent results, which is why the research team undertook this retrospective cohort analysis across two academic neonatal intensive care units.

The growth results were strikingly asymmetric. In the first month of life, infants receiving the soybean oil emulsion actually lost more weight z-score than those on the mixed emulsion, with a mean decline of 0.76 versus 0.59, a statistically significant difference. But from one month to 36 weeks postmenstrual age, the picture reversed dramatically. The soybean oil group gained length z-scores, improving by 0.13 on average, while the mixed emulsion group declined by 0.24. Over the full period from birth to 36 weeks, the soybean oil group also showed significantly better head circumference growth, with a post hoc mean difference of 0.152 favoring the older emulsion. Because head circumference is a proxy for brain growth, that difference drew particular attention.

Crucially, these growth findings survived statistical adjustment. In multivariable regression models controlling for gestational age, sex, intrauterine growth restriction, antenatal steroid exposure, length of stay, five-minute Apgar score, and average daily lipid dose, the mixed emulsion was associated with significantly worse changes in length z-score from one month to 36 weeks, and significantly worse head circumference z-score change from birth to 36 weeks. The early weight advantage of the mixed emulsion, however, lost statistical significance after adjustment, suggesting that the apparent early benefit may have been confounded by other factors, including the fact that infants in the mixed emulsion group received a slightly higher average daily lipid dose of 2.56 versus 2.42 grams per kilogram.

The morbidity data were equally provocative. Infants receiving the mixed emulsion had markedly higher rates of bronchopulmonary dysplasia, a chronic lung disease of prematurity, at 67.3% compared with 50.0% in the soybean oil group. After adjustment, the mixed emulsion conferred roughly 2.3-fold higher odds of the condition. Urinary tract infections followed a similar pattern, affecting 17.3% of the mixed emulsion group versus 8.4% of the soybean oil group, with an adjusted odds ratio near 2.0. Meanwhile, the two emulsions performed identically on the outcomes that motivated the newer blend’s development in the first place: rates of parenteral nutrition-associated cholestasis, hypertriglyceridemia, late-onset sepsis, necrotizing enterocolitis, retinopathy of prematurity, intestinal perforation, intraventricular hemorrhage, and mortality showed no significant differences between groups.

The authors point to mechanistic studies that may explain the pulmonary signal. In animal work, the mixed emulsion acted as a pro-oxidant and disrupted normal alveolar development through exaggerated programmed cell death. Clinical evidence has been similarly ambivalent about omega-3 supplementation: a recent meta-analysis found that enteral DHA supplementation did not protect against prematurity complications and in fact increased bronchopulmonary dysplasia risk, while other research showed that higher arachidonic acid levels in the first weeks of life correlated with reduced lung disease severity. The immune findings may also be biologically plausible, since omega-3 fatty acids dampen inflammatory responses, and fish oil supplementation has been shown to reduce the ability of monocytes to recruit neutrophils, potentially blunting the immune defense that keeps the urinary tract sterile.

Underlying all of this is the delicate balance between two rival fatty acids. DHA and arachidonic acid share common enzymatic pathways, and excessive DHA intake can suppress the body’s synthesis and tissue incorporation of arachidonic acid, which is essential for brain growth, cell membrane integrity, immune signaling, and bone formation. Animal studies have shown that high-DHA diets significantly reduce arachidonic acid concentrations in specific brain regions. The mixed emulsion, despite its more balanced profile, lowers the arachidonic acid to DHA ratio below the level observed in utero, and the optimal ratio for preterm infants has never been established. The authors suggest that this unresolved biochemical question may underlie the conflicting results across the literature, including earlier trials that found either no difference or advantages favoring the mixed emulsion.

The researchers are careful about the limits of their evidence. As a retrospective observational study, it could not control for every confounder, including antibiotic exposure, postnatal steroids, or the proportion of mother’s own milk in each infant’s diet, and the average lipid exposure of roughly 12 days was shorter than the interval over which some growth outcomes were measured. With multiple outcomes tested and no formal correction for multiple comparisons, false-positive findings remain possible, and the authors describe the growth results as hypothesis-generating rather than definitive. Still, their conclusion is pragmatic: until a well-designed randomized trial settles the question, the older soybean oil emulsion remains a reasonable, and possibly preferable, choice for preterm infants, a reminder that in neonatal nutrition, biochemical elegance does not guarantee clinical superiority.

Subject of Research: Comparison of soybean oil-based and mixed-oil intravenous lipid emulsions on growth and morbidity outcomes in preterm infants

Article Title: Comparison of growth and morbidity outcomes in preterm infants using two different intravenous lipid emulsions

Article References: Guy, A., Han, A., Kwiatkowski, D., Felmer, A. C., Gerber, C., Lee, R. H., Nayak, A., Goodman, R., Gothard, D., Ryan, R. M., Marasch, J., & Htun, Z. T. (2026). Comparison of growth and morbidity outcomes in preterm infants using two different intravenous lipid emulsions. Journal of Perinatology. https://doi.org/10.1038/s41372-026-02865-w

Image Credits: AI Generated

DOI: 10.1038/s41372-026-02865-w

Keywords: preterm infants, intravenous lipid emulsion, soybean oil, SMOF, parenteral nutrition, bronchopulmonary dysplasia, docosahexaenoic acid, arachidonic acid, growth z-scores, neonatology, urinary tract infection, omega-3 fatty acids