Intravenous lipid emulsions (ILE) play an important role in the parenteral nutrition (PN) of preterm infants by providing a concentrated source of essential fatty acids and energy (1, 2). Preterm infants often require higher doses of ILE than adults do to meet their growth, development, and daily metabolic needs (3). Infants who receive parenteral nutrition often experience a metabolic complication known as hypertriglyceridemia (HTG) (1, 3, 4). If left untreated, patients with HTG may develop complications, such as pancreatitis and lipid pneumonitis, and neurological changes, such as kernicterus (3, 5). In the past, neonatologists initiated treatment with conservative doses of macronutrients, including amino acids (AA) at 0.5 g/kg/day and ILE at 0.5 g/kg/day, as very low birth weight (VLBW) infants were believed to have limited metabolic tolerance for higher doses (6, 7). Following European Society for Pediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN) and American Society of Parenteral and Enteral Nutrition (ASPEN) guidelines, our current neonatal nutrition protocol initiates AA at 2.0 g/kg/day (2, 3, 8). In preterm infants with HTG, the ILE dose is typically reduced to 0.5–1.5 g/kg/day (3, 9). However, a subsequent reduction in ILE intake can lead to a decrease in overall energy intake, resulting in growth retardation, especially in VLBW infants (6). Another complication to consider when administering ILEs is fat overload syndrome (3, 10, 11). Fat overload syndrome manifests as hepatosplenomegaly, respiratory distress, jaundice, and spontaneous hemorrhage (3, 12). The side effects associated with fat overload syndrome are caused by elevated serum triglyceride (TG) levels, which occur when the rate of infusion exceeds the rate of lipid hydrolysis (3). Despite these side effects of HTG, there remains a lack of precise guidelines for adjusting lipid infusion rates based on TG, and there is a scarcity of studies analyzing the impact of HTG on preterm infants. ASPEN recommends the maximum rate for ILE is 0.15 g/kg/h for soybean oil, it depends on the ILE types (3). Regarding the 20% SMOFLipid® (Fresenius Kabi, Uppsala, Sweden) formula used in our hospital, the domestic approval requirements are as follows: ILE was gradually increased from 0.5 to 1.0 g/kg/day as daily fat to 3.0 g/kg/day, doses exceeding this are not recommended. The administration rate should not exceed 0.125 g/kg/h of fat (2). In the management of VLBW infants, it is crucial to provide adequate caloric intake through PN while minimizing complications that can arise from the use of ILE. Therefore, understanding the risk factors and significance of HTG is essential to ensure optimal care of these vulnerable infants.

The primary objective of this study is to analyze neonatal factors and time to normalize TG levels and to determine the effect on infants when ILE was reduced by 0–2 g/kg/day in the group with HTG levels, in single center study. In addition, we analyzed the threshold ILE infusion rate that could lead to HTG.

In preterm infants, ILE administration is a routine practice that supports caloric intake in those who do not receive sufficient EN. In this study, we analyzed the incidence and risk factors of HTG as well as the effects of ILE reduction and fat overload in preterm infants born before 32 weeks of gestation, following our nutrition guidelines.

HTG occurred in 37.7% of cases, which is higher than previous reported values (19.5%–29.7%) (1, 4, 13). Differences in diagnostic thresholds (e.g., 250 mg/dl vs. 200 mg/dl) and the inclusion of more immature and smaller infants in our study likely contributed to this discrepancy (1, 4, 13). Our cohort included infants with a mean GA of 27.3 ± 2.2 weeks and BW of 951.4 ± 239.8 g, lower than prior studies. The inclusion of a greater number of smaller and more premature infants in our study likely contributed to the higher incidence of HTG.

Consistent with existing research, BW was a significant risk factor for HTG, while GA and SGA were not (4, 6, 13, 14). This suggests that BW may better reflect nutritional vulnerability than GA alone. Our findings are supported by studies reporting that lower BW increases the risk of HTG, particularly in the early neonatal period when PN is heavily relied upon (5, 13). Reliance on PN may be further prolonged in cases requiring PDA treatment, due to delayed progression of EN (5). Previous studies have shown that early lipid infusion is not associated with an increased incidence of PDA (15–17). In cases where symptomatic PDA was treated with medication or ligation, the amount of enteral feeding likely decreased owing to the surgical conditions and side effects of the general condition. This likely increased their reliance on PN, a consideration supported by previous studies that reported similar results (6). Serum glucose levels were found to be significantly associated with HTG in our multivariate analysis, although the odds ratio was modest (OR = 1.01; 95% CI, 1.00–1.02). This may reflect complex metabolic interactions rather than a direct causal relationship. For example, the administration of ILE can increase free fatty acid levels, leading to reduced insulin sensitivity in peripheral tissues, impaired glycolysis, and increased gluconeogenesis—all of which may contribute to hyperglycemia in preterm infants (15, 18, 19). Furthermore, hyperglycemia in extremely low birth weight (ELBW) infants is common even in the absence of lipid infusion, due to factors such as stress, infections, steroid use, and insulin resistance (15, 20). Thus, while serum glucose level may serve as a marker of metabolic dysregulation associated with HTG, additional studies are needed to delineate the direction and mechanism of this association.

In this study, we analyzed the incidences of HTG, serum TG levels, and ILE infusion amounts based on GA and BW. A clear inverse trend was observed, where lower GA and BW were associated with higher HTG incidence, elevated TG levels, and greater lipid infusion doses. This reflects the limited enteral tolerance and higher dependence on PN among smaller and more immature infants. An exception to the trend was noted in the >2,000 g subgroup in Figure 1F, which showed a sudden increase in lipid dose. This is likely attributable to the small number of infants in this category and may not represent a true deviation. Interestingly, previous studies have reported an increasing trend in ILE dose with advancing GA (6). This discrepancy can be explained by the fact that in more mature infants, EN is typically established earlier, leading to a natural reduction in PN volume. In contrast, many ELBW infants in our cohort likely remained on PN during the first month due to feeding intolerance, resulting in higher cumulative lipid exposure.

In this study, with the exception of mortality, the incidence of other neonatal morbidities was higher in the HTG group (Table 1). Previous studies have investigated whether HTG can predict morbidity, and multivariate analysis revealed no significant association (4, 21, 22). However, other reports have indicated that HTG is a risk factor for ROP at low GA, specifically between 23 and 25 weeks (1). We investigated whether the time to TG normalization was associated with any neonatal morbidities and found a statistically significant association only with ROP stage 2 or higher. Docosahexaenoic acid (DHA) is an omega-3 fatty acid synthesized from alpha-linolenic acid, plays a crucial role in the development of the brain and retina (15, 23). Consequently, many studies have focused on the importance of early introduction of ILE in VLBW and ELBW infants (15). While Douglas et al. observed that the absence of ILE was associated with a higher risk of ROP (7), other reports did not find any significant difference in the ROP incidence between early and delayed ILE groups (15, 17, 24, 25). Given that ROP incidence is strongly associated with lower GA and BW, the observed association between HTG and ROP in our study could be influenced by these confounding factors. However, the persistence of the association with TG normalization time raises the possibility of a more direct metabolic link. One potential mechanism involves arachidonic acid metabolism. Arachidonic acid, an essential fatty acid in lipid metabolism, is a precursor to thromboxane—a compound produced via the cyclooxygenase pathway (26). ILE administration may alter this pathway and increase thromboxane levels, which can lead to retinal vasoconstriction, hypoxia, and reactive neovascularization, all of which are implicated in the pathogenesis of ROP (16, 26). Therefore, it is plausible that the faster normalization of TG levels in infants with HTG could mitigate excessive lipid metabolism and thromboxane production, thereby reducing the risk of ROP. While causality cannot be confirmed in our observational study, these findings suggest a potential biological link between lipid metabolism and retinal vascular pathology that merits further investigation.

We also examined whether reducing ILE dose could shorten the time to TG normalization. Although there was a negative correlation between the amount of lipid reduction and the number of days required for TG levels to normalize, this relationship was not statistically significant (β = −0.612; p = 0.556). Notably, higher TG levels were significantly associated with longer normalization times, suggesting that the severity of HTG itself is an important factor. This supports the clinical need to reduce the ILE dose when HTG occurs to facilitate recovery.

Finally, we determined the cutoff value for the ILE administration rate for all premature infants included in this study. This value was 0.332 cc/kg/h (1.6 g/kg/day), which was significantly lower than the domestic approval rate. With a sensitivity of 92.3% and an AUC of 0.711, this threshold may indicate an increased risk for HTG when exceeded. However, given the relatively low specificity (41.3%), this cutoff should not be interpreted as an absolute threshold for clinical decision making. Instead, it may serve as a reference point to highlight the importance of careful and individualized monitoring, particularly in preterm infants with lower birth weight or multiple risk factors. Our findings align with the ASPEN recommended dosing range of 1–3 g/kg/day, but suggest that HTG may still occur within this range, emphasizing the need for serum TG monitoring even when standard dosing is applied. As various factors influence HTG in preterm infants, caution is advised when approaching the upper limits of ILE infusion. To ensure adequate caloric support without inducing metabolic complications, lipid administration should be balanced with close TG monitoring and individualized risk assessment.

This retrospective study was subject to potential selection and reporting biases. Additionally, as this was a single-center study, it was limited by its small sample size. Another limitation is that while overall guidelines were followed for adjusting lipid administration, variations in the amounts adjusted by different individuals could introduce inconsistencies. However, this study had several strengths. It is the first to analyze the time required for HTG normalization while adjusting lipid administration and to examine the impact of faster normalization on neonatal morbidities. Additionally, this is the first study to analyze the critical ILE administration rate that can prevent fat overload syndrome, thereby providing valuable insights for clinical practice.

In conclusion, HTG is a common complication observed in premature infants. Lower BW, PDA, serum glucose levels, and higher ILE dosage were associated with increased HTG risk. Moreover, these findings suggest that while exceeding a certain lipid infusion rate may increase the risk of HTG, this value should not be interpreted as a fixed clinical threshold. Instead, it highlights the importance of individualized assessment and vigilant monitoring. Our results emphasize the need for customized and carefully titrated nutritional strategies to optimize growth while minimizing the metabolic risk in this vulnerable population. Furthermore, establishing precise guidelines for titrating ILE during HTG is necessary.