Lipid peroxidation assessment by malondialdehyde measurement in parenteral nutrition solutions for newborn infants: a pilot study JC Picaud 1,2 , JP Steghens 3 , C Auxenfans 4 , A Barbieux 4 , S Laborie 2 and O Claris 2 Neonatal Intensive Care Unit 1 , Arnaud de Villeneuve Hospital, University of Montpellier, Montpellier, France; Neonatal Intensive Care Unit 2 and Departments of Biochemistry 3 and Pharmacy 4 , Edouard Herriot Hospital, Claude Bernard University, Lyon, France Picaud JC, Steghens JP, Auxenfans C, Barbieux A, Laborie S, Reygrobellet B, Claris O. Lipid peroxidation assessment by malondialdehyde measurement in parenteral nutrition solutions for newborn infants: a pilot study. Acta Pædiatr 2004; 93: 241–245. Stockholm. ISSN 0803-5253 Aim: To determine malondialdehyde (MDA) concentrations in parenteral nutrition admixtures exposed to ambient room light, and in the serum of neonates. Methods: Using a new method to measure MDA specifically, this study analysed MDA of lipid-containing all-in-one admixtures provided by the pharmacy, with a composition identical to that used in routine clinical conditions. First, 12 admixtures were exposed to ambient light for 24 h, in the neonatal intensive care unit. Secondly, 18 solutions were either exposed to (n = 9) or protected from ambient light (n = 9) during the same period. Samples of admixtures were collected at baseline and 24 h later, for MDA measurement. Serum MDA was also randomly measured in orally fed healthy neonates. Results: After 24 h exposure to ambient room light, MDA concentrations in parenteral nutrition admixtures increased from 179 (129, 348) nmol l 1 to 5800 (1632, 14 679) nmol l 1 (p = 0.0002) [50th (10th, 90th) centiles]. When admixtures were protected from light, the increase in MDA was significantly lower than without protection: 187 (60, 429) nmol l 1 versus 13 696 (3472, 26 049) nmol l 1 (p = 0.0003). In 54 infants with a gestational age of 33 (28, 39) wk and a birthweight of 1750 (960, 3388) g, plasma MDA concentrations were 173 (98, 315) nmol l 1 . Conclusion: In solutions protected from light, MDA concentrations were low and were close to the serum MDA concentrations observed in orally fed neonates. Administration of all-in-one admixtures containing lipids in ambient lighting results in intravenous infusion of high levels of MDA which may present an additional source of morbidity in immature infants. This study confirms the need to protect parenteral admixtures from light. Key words: Lipid peroxidation, neonate, parenteral nutrition JC Picaud, Neonatal Intensive Care Unit, Arnaud de Villeneuve University Hospital, 371 Avenue Doyen Gaston Giraud, FR-34395 Montpellier, France (Tel.  33 4 67 33 66 09, fax.  33 4 67 33 62 28, e-mail. jc-picaud@chu-montpellier.fr) Total parenteral nutrition admixtures are useful for sick neonates, particularly for very low-birthweight infants. Complete admixtures for parenteral nutrition are com- posed of lipids, amino acids, glucose, electrolytes, vitamins and trace elements. Parenteral lipid emulsions provide essential fatty acids and are a concentrated source of calories for preterm infants. Lipid emulsions used in parenteral feeding of infants are susceptible to peroxidation, as judged indirectly by the formation of thiobarbituric acid reactive substances (TBARS), pen- tane and peroxides (1–4). Lipid peroxides and second- ary peroxidation products can be directly cytotoxic (5, 6). Free radicals have been implicated as mediators of the tissue injury that leads to bronchopulmonary dysplasia, retinopathy of prematurity and intraventri- cular haemorrhage in infants born preterm (7–9). Susceptibility to free radicals decreases with increasing gestational age (10). An association between elevated lipid peroxidation and poor outcome has been found in sick, very low-birthweight infants (11). As free radicals are short lived, they can be indirectly evaluated either by their primary reaction products with unsaturated fatty acids or by their secondary decom- position products such as hydroxypentenal, hydroxy- nonenal and malondialdehyde (MDA). MDA is highly cytotoxic because of its ability to bind proteins or nucleic acids very quickly (12). The determination of TBARS is a widely used approach for the evaluation of lipid peroxidation (4, 11–13). MDA measured as TBARS is, however, not specific for lipid peroxidation, because other aldehydes and non-lipid materials present in biological samples may also form thiobarbituric acid adducts (14). A new method to measure MDA specifically was recently described, which can be used for the determination of MDA in serum and in paren- teral nutrition solutions (15). In a preliminary study  2004 Taylor & Francis. ISSN 0803-5253 Acta Pñdiatr 93: 241±245. 2004 DOI 10.1080/08035250310022324