Dear Editor,

The article “Increased osmolality of breast milk with therapeutic additives” by L Srinivasan et al. reports the osmolality of supplements, drugs and fortifiers to breastmilk for preterm infants based on the assumption that laboratory measured osmolality presents an increased risk of necrotizing enterocolitis (NEC) for the preterm infant.[1] In the consideration of the association between hyperosmolar feedings and NEC there may be an important differentiation of osmolar substances into two groups: some substances create an osmotic gradient in vivo while others do not.[2] There is a difference between the osmolality that is measured in a laboratory and the effective osmolality in vivo.[3] Molecules that cross semi-permeable membranes, such as alcohol, do not present an osmotic load in vivo, but do contribute to measured osmolality by an osmometer (3). Molecules that do not set up an osmotic gradient are not likely to increase the risk of NEC by their osmolality.

Drugs as well as vitamin supplements often contain carrier molecules that can diffuse across membranes.[2] On the other hand, nutrients, such as sugars, amino acids, salts, minerals, and the constituents of breast milk fortifiers do not readily diffuse across membranes. Therefore, hyperosmolar feedings are more likely to pose an osmolar risk when the source of the hyperosmolality is from nutrients, but may not pose an osmolar risk when the source of the hyperosmolality is from medications and supplements.

The association between NEC and osmolality in preterm infants has been studied, inadvertently, in only one randomized clinical trial.[4] The remainder of the evidence that associates NEC with high osmolality feeds is from one quasi-experimental trial [5], three case reports [6-8] and some animal studies.[9-14] In all of these studies, the source of hyperosmolality associated with damage to the intestine was from substances that presented an osmolar load to the intestine.

Only those substances that need to be considered risky in terms of osmolality for potential inducement of NEC are those that create an osmotic gradient across the intestinal membranes in vivo. Some of the measured osmolality from drugs and vitamin supplements is not likely to present a risk from their osmolar load.

References

1. Srinivasan L, Bokiniec R, King C, Weaver G, Edwards AD. Increased osmolality of breast milk with therapeutic additives. Arch Dis Child Fetal Neonatal Ed 2004; 89(6):F514-F517.

2. Newmark HL, Johnson L, Etches PC, Carstensen JT. Osmolality-- limits of physical tests. Dev Pharmacol Ther 1988; 11(5):313-316.

3. Gennari FJ. Current concepts. Serum osmolality. Uses and limitations. N Engl J Med 1984; 310(2):102-105.

4. Book LS, Herbst JJ, Atherton SO, Jung AL. Necrotizing enterocolitis in low-birth-weight infants fed an elemental formula. J Pediatr 1975; 87(4):602-605.

5. Willis DM, Chabot J, Radde IC, Chance GW. Unsuspected hyperosmolality of oral solutions contributing to necrotizing enterocolitis in very-low-birth-weight infants. Pediatrics 1977; 60(4):535-538.

6. White KC, Harkavy KL. Hypertonic formula resulting from added oral medications. Am J Dis Child 1982; 136(10):931-933.

7. Tuladhar R, Daftary A, Patole SK, Whitehall JS. Oral gastrografin in neonates: a note of caution. Int J Clin Pract 1999; 53(7):565.

8. Wilcox DT, Fiorello AB, Glick PL. Hypovolemic shock and intestinal ischemia: a preventable complication of incomplete formula labeling. J Pediatr 1993; 122(1):103-104.

9. Norris HT. Response of the small intestine to the application of a hypertonic solution. Am J Pathol 1973; 73(3):747-764.

10. Stordahl A, Laerum F, Lunde OC, Aase S. The effects of water- soluble contrast media on luminal distension and blood flow in closed loops of small bowel in minipigs. Scand J Gastroenterol 1988; 23(8):991- 999.

11. Kameda H, Abei T, Nasrallah S, Iber FL. Functional and histological injury to intestinal mucosa produced by hypertonicity. Am J Physiol 1968; 214(5):1090-1095.

12. Torma MJ, Flannery MJ, French HG, Patressi GA, Gorman DL. Intestinal mucosal exfoliative response to graded hypertonic loads in neonatal and premature primates. Surg Forum 1978; 29:410-413.

13. Teichberg S, Lifshitz F, Pergolizzi R, Wapnir RA. Response of rat intestine to a hyperosmotic feeding. Pediatr Res 1978; 12(6):720-725.

14. Stordahl A, Haider T, Laerum F. Acute lethality after enteral administration of contrast media in anaesthetized rats with intestinal ischaemia. Acta Radiol 1989; 30(2):213-216.

Dear Editor,

The recommendations of maintaining enteral feeds of neonates of less than 400 mOsm/kg is clinically relevant in preventing necrotising enterocolitis (NEC).

It is reassuring that addition of fortifiers to expressed breast milk does not result in increased osmolality. The dilution chart proposed for supplements however has raised several issues:

1. A dilution volume of 5-18 mls seems unnecessary when neonates are receiving bolus enteral feeds of 40-75mls at any one period.
2. The more mature neonates receiving the nutritional supplements which greatly increase breast milk osmolality such as sodium ironedate syrup are at a lower risk of NEC.
3.The addition of substances such as chlorthiazide, spironolactone ,dexamethasone, ranitidine, domperidone, ompeprazole, gaviscon, phenytoin and phenobarbitone are clinically relevant to osmolality of breast milk inpatients receiving intensive care. These patients are also at the highest risk of NEC.

Extrapolating the osmolality of chloral hydrate and sodium acid phosphate to 5mls when they were added to 8 mls of breast milk flawed the methodology of the study. The involvement of 4 different centres to assess the osmolality of 7 nutritional additives also increases the operational bias and errors.

Repeat studies examining other additives, as mentioned above, which are used frequently in neonatal intensive care would be very beneficial . A subsequent dilution chart would also be a useful tool in administration of these drugs.