ABSTRACT
Background:
The last trimester of pregnancy is a period of rapid accretion of long‐chain polyunsaturated fatty acids, both in the central nervous system and the body as a whole. Human milk ...contains these fatty acids, whereas some preterm infant formulas do not. Infants fed formulas without these fatty acids have lower plasma and erythrocyte concentrations than infants fed human milk. Preclinical and clinical studies have demonstrated that single‐cell sources (algal and fungal) of long‐chain polyunsaturated fatty acids are bioavailable. A balanced addition of fatty acids from these oils to preterm formula results in blood fatty acid concentrations in low birth weight infants comparable to those of infants fed human milk.
Methods:
In the present study the growth, acceptance (overall incidence of discontinuation, reasons for discontinuation, overall incidence and type of individual adverse events), and plasma fatty acid concentrations were compared in three groups of infants fed a long‐chain polyunsaturated fatty acid‐supplemented preterm infant formula, an unsupplemented control formula, or human milk. The study was prospective, double‐blind (formula groups only), and randomized (formula groups only). Two hundred eighty‐eight infants were enrolled (supplemented formula group, n = 77; control formula group, n = 78; human milk group, n = 133).
Results:
Anthropometric measurements at enrollment, at first day of full oral feeding, and at both 40 and 48 weeks postconceptional age did not differ between the formula groups, whereas the human milk‐fed group initially grew at a lower rate. The incidence of severe adverse events was rare and not significantly different between formula groups. The groups fed either human milk or supplemented formula had long‐chain polyunsaturated fatty acid concentrations higher than those in the control formula group.
Conclusions:
The results of this study demonstrate the safety and efficacy of a preterm formula supplemented with long‐chain polyunsaturated fatty acids from single‐cell oils.
Linear random effects models for longitudinal data discussed by Laird and Ware (1982), Jennrich and Schluchter (1986), Lange and Laird (1989), and others are extended in a straight forward manner to ...nonlinear random effects models. This results in a simple computational approach which accommodates patterned covariance matrices and data insufficient for fitting each subject separately. The technique is demonstrated with an interesting medical data set, and a short, simple SAS PROC IML program based on the EM algorithm is presented.
PDF
