**Background:** Preterm infants born before 34 weeks gestation lack mature suck-swallow coordination and often require nasogastric tube (NGT) feeding. Previous research has shown that tube feeding can reduce the fat content of human milk due to adherence to tubing, but little is known about its effect on infant formula. Given that formula is a primary nutrition source for many very low birth weight (VLBW) infants, any loss of macronutrients could have clinical consequences for growth and development. This study aimed to evaluate changes in macronutrient content of various infant formulas after passage through a feeding tube under simulated real-life conditions.
**Methods:** Eleven infant formulas commonly used in the NICU at Tel Aviv Sourasky Medical Center were selected: 2 preterm formulas (Nos. 1 and 2), 4 extensively hydrolyzed formulas (Nos. 3–6), 2 amino acid–based formulas (Nos. 7 and 8), and 3 standard cow’s milk–based formulas for term infants (Nos. 9–11). Three formulas were ready-to-feed (RTF) and the others were powder-based, prepared per manufacturer guidelines. To mimic clinical practice, the average feeding time for a 5-ml human milk sample was first measured in 22 preterm infants (gestational age 26–34 weeks), yielding a mean of 14 minutes 30 seconds. Simulated bolus feeding was then performed using an infusion pump (Alaris) connected to a feeding tube (Metric/x-ray, 40 cm CH 05) to transfer 30 ml of formula at 20 ml/h into collection tubes. For each formula, 10 samples were analyzed before and after passage through the tube. The tube was changed after each simulated meal and not washed. Macronutrient content (fat, protein, carbohydrate, energy) was measured using a MIRIS milk analyzer based on mid-infrared transmission spectroscopy. Paired t-tests compared pre- and post-infusion values; stepwise multiple regression assessed the effect of formula type. A total of 220 measurements were performed, but 20 measurements from one amino acid–based formula (No. 8) were excluded due to technical error.
**Key Results:** Variations in at least one macronutrient were observed in 5 out of 10 formulas (Nos. 1, 2, 3, 5, and 6). Specifically:
- Formula No. 1 (preterm, RTF): Fat decreased from 5.98 ± 0.09 to 5.85 ± 0.08 g/dL (2% loss, p=0.002); energy decreased from 92.8 ± 0.9 to 91.4 ± 1.3 kcal/dL (1.5% loss, p=0.007). Protein and carbohydrate changes were not significant.
- Formula No. 2 (preterm, RTF): Protein decreased from 1.58 ± 0.04 to 1.49 ± 0.03 g/dL (2.2% loss, p=0.005); energy decreased from 67.6 ± 0.69 to 66.8 ± 0.63 kcal/dL (1% loss, p<0.001). Fat and carbohydrate changes were not significant.
- Formula No. 3 (extensively hydrolyzed): Fat decreased from 4.32 ± 0.06 to 4.25 ± 0.08 g/dL (1.6% loss, p=0.04).
- Formula No. 5 (extensively hydrolyzed): Fat decreased from 5.2 ± 0.04 to 5.1 ± 0.06 g/dL (1.5% loss, p=0.01).
- Formula No. 6 (extensively hydrolyzed): Carbohydrate decreased from 6.86 ± 0.1 to 6.64 ± 0.04 g/dL (3.3% loss, p=0.002).
No significant changes were found in formulas Nos. 4, 7, 9, 10, or 11. Stepwise multiple regression showed that formula type significantly affected post-infusion fat, protein, and energy content (p<0.001 for all).
**Clinical Implications:** This study demonstrates that even at a slow bolus feeding rate (20 ml/h), passage through a feeding tube can cause small but statistically significant losses of fat, protein, or carbohydrates in some infant formulas. The losses were modest (1–3.3%) but could be clinically relevant for VLBW infants who rely on precise nutrient delivery for growth and neurodevelopment. Fat loss is particularly concerning because lipids are the primary energy source and provide essential fatty acids critical for brain and retinal development. The variability among formulas suggests that formulation differences (e.g., fat composition, protein type) influence adherence to tubing. The authors note that previous studies found no significant losses with formula, possibly due to different infusion rates or formula compositions. Limitations include the lack of direct biochemical analysis (using near-infrared spectroscopy instead) and absence of clinical outcome data (e.g., weight gain). Larger studies using direct biochemistry and examining different feeding practices are needed to confirm these findings and determine whether formula concentration or volume adjustments are warranted to compensate for nutrient losses during tube feeding.