Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Total antioxidant capacity, vitamin A and E, and fatty acid content of human milk Tijerina Saenz, Alexandra 2007

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2007-0616.pdf [ 5.81MB ]
Metadata
JSON: 831-1.0101135.json
JSON-LD: 831-1.0101135-ld.json
RDF/XML (Pretty): 831-1.0101135-rdf.xml
RDF/JSON: 831-1.0101135-rdf.json
Turtle: 831-1.0101135-turtle.txt
N-Triples: 831-1.0101135-rdf-ntriples.txt
Original Record: 831-1.0101135-source.json
Full Text
831-1.0101135-fulltext.txt
Citation
831-1.0101135.ris

Full Text

TOTAL ANTIOXIDANT CAPACITY, VITAMIN A A N D E, A N D FATTY ACID CONTENT OF H U M A N M I L K by Alexandra Tijerina Saenz BSc. Tecnologico de Monterrey, 2003 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L M E N T OF T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF M A S T E R OF S C I E N C E in The Faculty of Graduate Studies (Human Nutrition) T H E U N I V E R S I T Y OF BRIT ISH C O L U M B I A October 2007 © Alexandra Tijerina Saenz, 2007 ABSTRACT Tota l antioxidant capacity ( T A C ) o f human m i l k was used as a measurement to evaluate m i l k oxida t ive stabil i ty. The O x y g e n R a d i c a l ' A b s o r b a n c e Capac i ty ( O R A C ) assay was used to measure oxida t ive stability o f human m i l k based o n the advantages o f the method, the photostabi l i ty o f the fluorescent probe f luorescein ( F L ) , and the specifici ty for antioxidant ac t iv i ty against pe roxy l radicals. The present study consisted o f two sections; firstly to standardize and validate the O R A C F L assay for use i n determining T A C o f human m i l k and second to determine potential relationships between different nutri t ional components and T A C levels o f mature human m i l k . The O R A C F L assay was standardized and val idated by determining the linearity, precis ion and accuracy o f the assay and by determining an op t imal d i lu t ion for human mi l k . G o o d l ineari ty (R^>0.990) and prec is ion (2.2%) were achieved and the assay accuracy was 95.4%, w h i c h was obtained by the recovery o f T r o l o x standard after experimental analysis. S i x different m i l k d i lu t ions were used to evaluate m i l k T A C and a d i lu t ion o f 200x was chosen for routine O R A C F L analyses. A convenience sample consis t ing o f human m i l k samples (n=ll) was obtained from Canadian w o m e n at mon th 1 postnatally. Contents o f m i l k fatty acids ( F A ) , v i t amin A (al l -fr-ara'-retinol), v i t a m i n E (a-tocopherol , 8-tocopherol, and y-tocopherol) , and T A C were measured and the b i o l o g i c a l act ivi ty o f v i t amin E and a total fatty ac id unsaturation index (UI) were calculated. A total o f 60 m i l k samples were used i n the f inal analyses, due to the presence o f 17 samples w i t h T A C values be low detectable levels . A c c o r d i n g to the results, T A C o f mature human m i l k was pos i t ive ly attributed to m i l k a-tocopherol (r= 0.439, p<0.001). H i g h e r contents o f a-tocopherol s ignif icant ly increased m i l k oxidat ive stability. O n the other hand, different contents o f v i t amin A or fatty acids or U I had no effect on m i l k T A C . These results suggested that oxidat ive stabili ty o f mature human m i l k cou ld reflect the complex sum o f different antioxidants and/or mechanisms, but conf i rmed the importance o f the presence o f antioxidants, such as v i t amin E , i n protecting against free radical- induced l i p i d peroxida t ion reactions. i i TABLE OF CONTENTS A B S T R A C T ii L IST OF T A B L E S vi i L IST OF F I G U R E S .... ix L IST OF A B B R E V I A T I O N S x i A C K N O W L E D G E M E N T S xi i i G E N E R A L I N T R O D U C T I O N .'. 1 C H A P T E R 1: L I T E R A T U R E R E V I E W 5 1.1 Human M i l k 5 1.1.1 Colostrum, Transitional, and Mature M i l k 5 1.1.2 Nutritional Composition of Mature Human M i l k 6 1.1.3 Human M i l k Protein 6 1.1.4 Human M i l k Carbohydrates 8 1.1.5 Human M i l k Vitamins 8 1.1.6 Human M i l k Lipids 9 1.2 Fatty Acids and Neonatal Health 11 1.3 Antioxidant System in Human M i l k and Neonatal Health 14 1.3.1 Total Antioxidant Capacity of Human M i l k 14 1.3.2 Vi tamin C and Neonatal Health 15 1.3.3 Vi tamin A and Neonatal Health 16 1.3.4 Vi tamin E and Neonatal Health 17 1.4 Oxidative Stress and Neonatal Health 19 1.5 Oxidative Stability of Human Mi l k : The Role of R O S and L ip id Peroxidation ... 20 1.6 Measures of Oxidative Status 22 1.6.1 D N A oxidation 22 1.6.2 Protein oxidation 22 1.6.3 L ip id oxidation 23 1.7 Measures of Total Antioxidant Capacity 24 i i i 1.7.1 D P P H Radical Scavenging Activity 24 1.7.2 Ferric Reducing Antioxidant Power (FRAP) 24 1.7.3 Trolox Equivalent Antioxidant Capacity ( T E A C ) 24 1.7.4 Total Radical Trapping Parameter (TRAP) 25 1.7.5 Oxygen Radical Absorbance Capacity ( O R A C ) 25 C H A P T E R 2: S T A N D A R D I Z A T I O N OF O R A C F L A S S A Y F O R H U M A N M I L K 27 2.1 I N T R O D U C T I O N 27 2.2 M A T E R I A L S A N D M E T H O D S 28 2.2.1 Human M i l k Samples 28 2.2.2 Materials and Equipment 28 2.2.3 Standardization of the O R A C F L Assay 28 2.2.3a The O R A C F L Assay 28 2.2.3b Linearity and Precision of the O R A C F L Assay 30 2.2.3c Human M i l k Dilution 30 2.2.3d Antioxidant Recovery Experiments 31 2.2.3e Antioxidant Capacity of Human M i l k 32 2.2.4 Data Analysis 33 2.3 R E S U L T S 34 2.3.1 The O R A C F L Assay 34 2.3.2 Linearity and Precision of the O R A C F L Assay 35 2.3.3 Human M i l k Dilution 36 2.3.4 Antioxidant Recovery Experiments 38 2.3.5 Antioxidant Capacity of Human M i l k 40 2.4 D I S C U S S I O N 42 2.5 C O N C L U S I O N 44 C H A P T E R 3: A S T U D Y O N T H E T O T A L A N T I O X I D A N T C A P A C I T Y , V I T A M I N A A N D E A N D F A T T Y A C I D C O N T E N T OF H U M A N M I L K 45 3.1 I N T R O D U C T I O N 45 3.2 M A T E R I A L A N D M E T H O D S 46 3.2.1 Human M i l k Samples 46 3.2.2 Collection of M i l k Samples 46 3.2.3 Background of Participants 46 3.2.4 Demographics of Participants 47 3.2.5 Gestational Dietary Intakes of Participants 47 3.2.6 Chemical Analyses of Human M i l k 47 3.2.6a Fatty A c i d Analysis 48 3.2.6b Vi tamin A and E Analysis 48 3.2.6c Total Antioxidant Capacity Analysis 49 3.3 D A T A A N A L Y S I S 49 3.3.1 Screening M i l k Samples 50 3.3.2 Evaluation of M i l k T A C , Vitamins, and Fatty Ac i d Composition 50 3.3.3 Evaluation of Human M i l k Oxidative Stability 50 3.3.4 Statistical Analysis 50 3.4 R E S U L T S 51 3.4.1 Demographic Characteristics 51 3.4.2 Maternal Dietary Intakes during Gestation 52 3.4.3 Antioxidant Capacity, Vitamin, and Fatty Acids of Human M i l k 54 3.4.4 Oxidative Stability of Human M i l k 58 3.5 D I S C U S S I O N 63 3.5.1 Participants: Background, Demographics, and Dietary Intakes 63 3.5.2 Antioxidant Capacity, Vitamin, and Fatty Ac ids of Human M i l k 64 3.5.3 Oxidative Stability of Human M i l k 68 3.6 C O N C L U S I O N 69 G E N E R A L C O N C L U S I O N 70 R E F E R E N C E S 72 A P P E N D I X A : M A T E R N A L D I E T A R Y I N T A K E S 92 A . 1 Maternal Dietary Intakes 92 A.2 Maternal Supplement Intakes 92 v A.3 Data Analysis 92 A.3.1 Dietary Intakes using the E S H A Food Processor 92 A.3.2 Calculating Vi tamin A Intake 93 A3.3 Calculating Vi tamin E Intake 94 A.3.4 Calculating the Acceptable Macronutrient Distribution Range ( A M D R ) 94 A . 4 Statistical Analyses 95 A P P E N D I X B: H P L C M E T H O D F O R D E T E R M I N A T I O N OF V I T A M I N S A A N D E IN H U M A N M I L K 96 B. l Human M i l k Samples 96 B.2 Chemicals and Equipment 96 B.3 M i l k Fat Extraction 96 B.4 Determination of H P L C Method Parameters 97 B.5 Vi tamin Concentration in M i l k Samples 99 B.6 Linearity of Calibration Curves of Vitamin Standards 99 B. 7 Recovery 102 A P P E N D I X C : D A T A A N A L Y S I S , F O R M U L A S , A N D S T A T I S T I C A L R E S U L T S .... 104 C l Box Plots 104 C . l . l Interpretation of Box Plots 104 C.1.2 Comparison of milk D H A contents 105 C. 2 Antioxidant Capacity, Vitamin, and Fatty Acids of Human M i l k 105 C.2.1 Mult iple Regressions 105 C.2.2 Bivariate Correlations 108 C.2.3 Plots of Bivariate Correlations 109 C.3 Oxidative Stability of Human M i l k I l l C.3.1 Calculating Tertiles I l l C.3.2 Calculating Odds Ratio (OR) I l l C.3.3 Interpreting Odds Ratio (OR) 112 C.3.4 Calculating Unsaturation Index (Ul) 113 LIST OF TABLES Table 1.1 Nutritional components in colostrum, transitional, and mature human milk 5 Table 1.2 Nutritional components in human milk, bovine milk, and infant formula 6 Table 1.3 Comparison of protein distribution of mature human milk and bovine milk 7 Table 1.4 Fatty acid composition of mature human milk, bovine milk, and infant formula.. 10 Table 2.1 Specific chemicals needed to make Trolox Calibrator in O R A C F L assay 29 Table 2.2 Chemicals needed to analyze Trolox Calibrator and human milk samples 30 Table 2.3 Composition of spiked human milk samples 31 Table 2.4 Composition of milk spiked with known concentration of Trolox standard 32 Table 2.5 Composition of Trolox Calibrator spiked with a known volume of human milk... 33 Table 2.6 Results from linear regression analysis of Trolox Calibration 35 Table 2.7 Average net A U C values of Trolox Calibrator (n=3) 36 Table 2.8 Average net A U C values and R of different human milk dilutions 37 Table 2.9 Accuracy on experimental calculation of concentration of Trolox added to human milk 39 Table 2.10 Accuracy on experimental calculation of concentration of Trolox added to a human milk sample 40 Table 3.1 Demographic characteristics of participants («=60) 51 Table 3.2 Maternal average daily intakes of macronutrients given as percent of energy intake during gestation («= 29) 52 Table 3.3 Maternal daily intakes for individual fatty acids during gestation (n= 29) 53 Table 3.4 Maternal average daily intake of antioxidant vitamins during gestation (n= 29)... 53 Table 3.5 T A C and vitamin content in mature human milk (n-60) 54 Table 3.6 Fatty acid content and UI in mature human milk (g/lOOg FA) («=60) 55 vii Table 3.7 Tertiles of low (I), medium (II), and high (III) contents of mi lk variables 59 Table 3.8 Odds ratio and 95% CI for having low levels o f mi lk T A C value 61 Table 3.9 Odds ratio and 95% CI for having high levels of mi lk T A C value 62 Table A . l Conversion factor for food sources 93 Table A.2 Conversion factor for supplemental vitamin A 94 Table B . l H P L C parameters for determination of vitamin A and E isomers 98 Table B.2 Regression equations and R 2 of calibration curves for vitamins A and E 101 Table B.3 Composition of mixtures used for recovery analyses 102 Table B.4 Recovery o f vitamin isomers added to mi lk samples A and B 103 Table C l Summary of model 1, showing R 2 105 Table C.2 Significance of the model 105 Table C.3 Beta coefficients and significance of independent variables 106 Table C.4 Summary of model 2, showing R 2 107 Table C.5 Significance of model 2 107 Table C.6 Beta coefficients and significance of independent variables 107 Table C.7 Bivariate relationships, Pearson correlation coefficients, and significance 108 Table C.8 Tabular formulation of odds ratio (OR) 112 LIST OF FIGURES Figure 1.1 Chemical structure of the essential fatty acids omega-3 a-linolenic acid (A) and omega-6 linoleic acid (B) 11 Figure 1.2 Scheme of omega-6 and omega-3 fatty acid desaturation and elongation 12 Figure 1.3 Strucure of L- ascorbic acid and L-dehydroascorbic acid 15 Figure 1.4 Structure of all trans-xoXmol 16 Figure 1.5 Structure of a-tocopherol 18 Figure 1.6 Scheme of the antioxidant activity of a-tocopherol against peroxyl radicals.... 18 Figure 1.7 Scheme of production of reactive oxygen species (ROS) in the cell 20 Figure 1.8 Scheme of reactions in l ipid peroxidation 21 Figure 2.1 Effect of concentration of Trolox on Trolox Calibrator decay curve 34 Figure 2.2 Standard curve for Trolox Calibrator after data transformation 34 Figure 2.3 Net A U C values of human milk and concentration at dilutions 200x and 225x.. 37 Figure 2.4 Net A U C vs milk concentration of human milk spiked with Trolox standard... 38 Figure 2.5 Net A U C values of non-spiked and Trolox spiked human milk 39 Figure 2.6 Increase on net A U C values while adding human milk to Trolox Calibrator .... 41 Figure 3.1 Relationship between milk T A C and a-tocopherol («=77), showing samples with undetectable T A C values 49 Figure 3.2 Relationship between milk T A C and a-tocopherol 56 Figure 3.3 Scatter plot to show milk T A C and all-fr-cms-retinol 57 Figure 3.4 Scatter plot to show milk D H A and milk T A C 57 Figure 3.5 Scatter plot to show milk UI and milk T A C 58 Figure B. 1 Chromatogram corresponding to standards (mixed) 97 Figure B.2 Chromatogram corresponding to a milk sample 98 Figure B.3 Calibration curve for a-tocopherol 100 Figure B.4 Calibration curve for 8-tocopherol 100 Figure B.5 Calibration curve for y-tocopherol 100 Figure B.6 Calibration curve for all-fram ,-retinol 101 Figure C l Comparison of box plot with normal distribution curve 104 Figure C.2 Box plot comparing milk D H A contents at month 1 postnatally (n=60) 105 Figure C 3 Relationship between milk a-tocopherol and alWra/zy-retinol contents 109 Figure C.4 Relationship between milk D H A and E P A contents 109 Figure C.5 Relationship between milk D H A and A L A contents 110 Figure C.6 Relationship between milk L A and A L A 110 Figure C.7 Standardized distribution curve, showing tertiles areas and z-scores I l l x LIST OF ABBREVATIONS ABTS 2,2'-azinobis(3-ethylbenzothiazoline 6-sulfonate) A A P H 2,2'-azobis(2-amidinopropane) dihydrochloride DPPH 2,2-diphenyl-1 -picrylhydrazy DCFH 2' ,7' -dichlorodihydro fluorescein 8-OHdG 8-hydroxy-2'-deoxyguanosine A M D R Acceptable Macronutrient Distribution Range A A Arachidonic acid A U C Area under the curve A U C N s A U C value of non-spiked milk sample A U C S A U C values of the spiked sample A U C J C A U C values of the Trolox Calibrator A U C T E Experimental A U C values of 20uM Trolox added B-PE B-phycoercythrin CAT Catalase CFRI Child & Family Research Institute C V Coefficient of variation %cv Coefficient of variation CI Confidence interval CD Conjugated dienes R 2 Determination coefficient DRI Dietary Reference Intake D A D Diode-array detector D H A Docosahexaenoic acid EPA Eicosapentaenoic acid F A Fatty acids FRAP Ferric Reducing Antioxidant Power FOX-2 Ferrous oxidation assay QI First quartile FL Fluorescein FSS Fluorescein sodium salt FFQ Food Frequency Questionnaire FFA Free fatty acids G L C Gas-liquid chromatography GPx Glutathione peroxidase HSPs Heat shock proteins H 2 0 2 Hydrogen peroxide OH» Hydroxyl radical I S S F A L International Society for the Study of Fatty Acids and Lipids. IU International Units IQR Interquartile range L A Linoleic acid A L A Linolenic acid H P L C L iquid chromatography L C - P U F A Long chain polyunsaturated fatty acids L D L L o w density lipoprotein M D A Malondialdehyde M U F A Monounsaturated fat O R Odds ratio co-3 Omega-3 fatty acids co-6 Omega-6 fatty acids O R A C Oxygen Radical Absorbance Capacity O R A C B - P E Oxygen Radical Absorbance Capacity using B-phycoercythrin O R A C F L Oxygen Radical Absorbance Capacity using fluorescein P B Phosphate buffer P U F A Polyunsaturated fatty acids R O S Reactive oxygen species R D A Recommended Dietary Allowance R A E Retinol Activi ty Equivalents R E Retinol Equivalent R -PE R-phycoerythrin S F A Saturated fat ±SD Standard deviation S E M Standard error of the mean SOD Superoxide dismutase 0 2 ' - Superoxide radical T B A R S Thiobarbituric acid reactive substances Q3 Third quartile a-TTP Tocopherol binding protein T E Tocopherol Equivalents T A C Total Antioxidant Capacity T R A P Total Radical Trapping Parameter T E A C Trolox Equivalent Antioxidant Capacity U l Unsaturation index U L Upper Level x i i ACKNOWLEDGEMENTS I would like to thank to all those who have helped me from the process to become a graduate student at the University of British Columbia to the end of this important stage of my life. I want to thank the Department of Food, Nutrition and Health from the Faculty of Land and Food Systems for giving me the opportunity to perform necessary research experiments. In addition, this research study could not be possible without the help of B C Women's M i l k Bank and participants who donated human milk. I am deeply indebted to my supervisor Dr. David Kitts, Director and Professor from the Department of Food, Nutrition and Health, who gave me the opportunity to work for his research team and who supervised my studies and research project, and helped me in all the time of research and writing of this thesis. I have furthermore to thank Dr. Sheila Innis who gave me the opportunity to work in my research project by a partnership with the Chi ld and Family Research Institute, Vancouver B.C. , and supervised the progress of my thesis research. M y professors, colleagues and technicians from the Programs of Food Science and Human Nutrition supported me in my research work. I want to thank them for their help, support, interest and valuable hints. Gratitude to Dr. Zhaoming X u , Dr. Susan Barr, Dr. Eunice L i -Chan, Dr. Christine Seaman, Dr. Charles Hu, Russell Friesen, Janette K., Dr. Parastoo Yaghmaee, X ium in Chen, Ingrid El is ia, Valerie Skura, Dr. Pedro Aloise and Tram Nguyen. Thanks to my advisor from the Department of Statistics, Al ine Tabet, who helped me in data analysis and the understanding of statistics. I would like to thank all people who have believed in me from the beginning of this journey; those people who believed in me during the process; but mostly, a special Thank You to all those who still believe in me, especially my family and friends. Especially, I would like to give my special thanks to my parents Saul and Ensueno and to my brothers Saul and David whose patient love and encouragement enabled me to complete this work. And thank Y O U for always being with me, my provision, my guidance, my strength. Xlll G E N E R A L I N T R O D U C T I O N Human milk represents the ideal food source that provides the breastfed infant with nutritional and bioactive factors, distinct from bovine milk and infant formulas. Human milk is digested twice as fast as bovine milk or formula (Newton, 2004) and is attributed to different unique factors that include protein composition, casein distribution, and relative smaller micelle size of milk containing proteins. As a result, digestion of human milk produces a finer, softer coagulate compared to digested bovine milk or formula (Malacarne et al., 2002). Some examples of human milk factors include normally present and digestion derived bioactive peptides, such as lactoferrin, lysozyme and casein phosphopeptide that provide antioxidant protection (Worthington-Roberts, 1997; Picciano, 1998; Lindmark-Mansson and Akesson, 2000; Hamosh, 2001; Morrow et al., 2004; Newton, 2004; Kitts, 2005, Wong et al., 2006), complex oligosaccharides that enhance host defenses in the breastfed infant (Hamosh, 2001), and the presence of long-chain polyunsaturated fatty acids (LC-PUFA) that are essential for visual acuity, neural development, neurotransmitter metabolism, and behavior of the infant (Innis, 2003; Silva et al., 2005; Helland et al., 2006; Hadders-Algra et al., 2007; Watkins et al., 2007). Different studies have reported that maternal fatty acid intake during gestation and lactation may provide positive influences regarding the fatty acid composition in human milk (Francois et al, 1998; Turoli et al., 2004; Agostoni, 2005; Xiang et al., 2005). One concern is the susceptibility of unsaturated fatty acids present in human milk to lipid oxidation, which is caused by the attack of free radicals, known as reactive oxygen species (ROS), on the methylene carbon located in between two double bonds of the polyunsaturated fatty acids (PUFA) (Schneider, 2005). Lipid oxidation of human milk not only results in off-flavor formation that coincides with a typical rancid odor, but also the presence of oxidized and potentially toxic compounds, such as free radicals, that are ingested by breastfed infants (Turoli et al., 2004). Increased levels of LC-PUFA in biological systems require increased levels of antioxidant compounds to prevent lipid oxidation. For example, alpha-tocopherol has been reported to be the main antioxidant compound present to neutralize peroxyl radicals generated in biological lipid systems (Kitts, 1997; Ternay and 1 Sorokin, 1997; Oostenbrug et al., 1998; Lindmark-Mansson and Akesson, 2000; Schneider, 2005). The chosen nutritious food source (human or bovine milk or formula) can affect the overall health of the neonate or the infant and must contribute to the antioxidant defense system against R O S (Turoli et al., 2004). Thus, it is imperative that mother's milk, a natural nutritious food for the growing neonate, contain natural stabilizers, such as antioxidants and sequestering agents that ensure a stable l ipid component that wi l l retain essential nutrients that are otherwise labile, and reduce exposure of the breastfed infant to exogenous sources of oxidative stress. This is particularly crucial to the health of the neonate who does not have a mature and fully functional antioxidant system. Interestingly, human milk contains an antioxidant system which maintains the oxidative stability of mi lk and thus protects the breastfed infant from the potential negative health effects derived from exposure to R O S . The human milk antioxidant system is comprised of bioactive components found to have antioxidant activity such as vitamins A , E, and carotenes, bioactive peptides like lactoferrin and lysozyme, and the enzymes catalase, superoxide dismutase, and glutathione peroxidase (L 'Abbe and Friel , 2000; Lindmark-Mansson and Akesson, 2000; Hamosh, 2001; Shoji et al., 2004; Kasapovic et al., 2005; Shoji et al., 2005). Protection against l ipid oxidation by antioxidant enzymes present in human milk has been evaluated. Addit ion of antioxidant enzymes to infant formula resulted in an increase protection against l ipid oxidation of formula (Friel et al., 2002). Another study has reported that human milk protects the gastrointestinal tract of the breastfed infant against R O S oxidative damage (Shoji et al., 2005). The complexity of the antioxidant system and possible synergistic interactions among the different antioxidant compounds present, have resulted in the development of assays and methodologies that measure total antioxidant capacity (TAC) . In addition, measurement of a single antioxidant or measurement of each antioxidant separately wi l l be less representative than measures of total antioxidant status. The T A C of human mi lk has been evaluated and compared to that of infant formulas. A higher antioxidant protection against R O S was found in human milk than in formula (Friel et al., 2002; Aycicek et al., 2006). A decrease in total antioxidant activity in human milk was also reported with different storage conditions (Hanna et al., 2004; Miranda et al., 2004; Turoli, 2004). The Oxygen Radical Absorbance Capacity ( O R A C ) assay has been developed and validated to measure the antioxidant 2 capacity of numerous foods, dietary supplements, juices and wines, and biological samples such as plasma and urine (Prior et al., 2003; Davalos et al., 2004; W u et al., 2004; Fernandez-Pachon, et al., 2005). This methodology has also been used to study the antioxidant capacity of colostrum, transitional, and mature human milk of Italian women (Alberti-Fidanza et al., 2002); however, there is no indication that the O R A C assay was validated for the measurement of human milk T A C . The purpose of this research was to standardize and validate the improved O R A C F L assay, which uses fluorescein (FL) as the fluorescent probe, to generate quantitative measures of the oxidative stability of human milk, and to determine potential effects of different milk components on the milk T A C . For the standardization and validation of the O R A C F L assay, human mi lk samples were provided by the B C Women's M i l k Bank. For the second purpose, mature human milk was obtained at first month postpartum and experiments were initiated to quantitate the amounts of antioxidant vitamins, such as all-trans-rctinol, and vitamin E (a, 8, and y-tocopherols). The content of individual free fatty acids and total degree of fatty acid unsaturation were also measured and were related to mi lk T A C values. The hypotheses of this research were: 1. The O R A C F L assay represents a reliable measurement of the total antioxidant capacity (TAC) of human milk. 2. A general relationship exists between milk T A C measurements and the presence of major l ipid soluble components (e.g. fat soluble vitamins and fatty acids) 3a. The oxidative stability of mature human milk, measured as milk T A C , is greater in individuals that have high antioxidant vitamin content in milk. 3b. The oxidative stability of human milk can be adversely affected in mothers that have elevated milk free fatty acid contents. 3 The research objectives were: 1. To standardize the O R A C F L assay for human milk and validate the procedure for linearity, precision, and accuracy required for the assessment of the oxidative stability of human milk. 2. To measure the lipid-soluble antioxidant vitamin contents (all-toms-retinol and a, 8, and y-tocopherols), of mature human milk using liquid chromatography. 3. To determine possible relationships between specific human milk components and milk T A C . 4. To characterize the oxidative stability of human milk in terms of specific levels of antioxidant vitamins and fatty acid contents. 4 CHAPTER 1: LITERATURE REVIEW 1.1 Human Milk Human milk is considered the ideal food to meet optimal nutritional and immunological requirements for the growth and development of newborn infants at least during the first six months of life (Health Canada, 1999; Newton, 2004). It is a unique food not only for its source of energy and nutritional components, but also for the presence of bioactive agents that are distinct from bovine milk and infant formula (Lawrence, 1994; Picciano, 1998; Malacarne et al., 2002; Newton, 2004). 1.1.1 Colostrum, Transitional, and Mature Milk Nutritional composition of colostrum, transitional, and mature human milk is shown in Table 1.1. The first week of postpartum mammary secretion consists of a yellowish and thick fluid known as colostrum. Protein, fat-soluble vitamins, and minerals are present in higher levels in colostrum than in transitional or mature milk (Lawrence, 1994) and it is a source of immune and antioxidant components for the breastfed newborn (Kasapovic et al., 2005). Table 1.1 Nutritional components in colostrum, transitional, and mature human milk Colostrum Transitional Milk Mature Milk Energy (Kcal/dL) 65 65 67 Total Protein (g/dL) 2.3 - 0 . 9 - 1 . 7 Whey : Casein ratio 90 : 10 - 6 0 : 4 0 Total Fat (g/dL) 2.9 3.6 3.8 Lactose (g/dL) 5.3 - 6.8 - 7.3 Vitamin C (mg/L) 59 71 5 0 - 1 0 0 Vitamin A (ug/L) 1500 -2000 880 2 0 0 - 4 0 0 Vitamin E (mg/L) 8 - 1 5 9 2 . 5 - 4 . 0 (Lawrence, 1994; Picciano, 2001; Newton, 2004) 5 The transitional mi lk phase lasts from day 7-10 postpartum to 2 weeks postpartum. The phase of mature milk starts after 2 weeks postpartum (Lawrence, 1994). The concentration of immunoglobulins, total protein, and fat-soluble vitamins and carotenoids decreases with the stage of lactation; while the content of lactose, fat, and water-soluble vitamins increases (Lawrence, 1994; Picciano, 1998; Newton, 2004). 1.1.2 Nutritional Composition of Mature Human Milk The nutritional composition of human milk is distinctive from bovine milk and infant formula as shown in Table 1.2. Human milk contains less protein, similar fat, and has a relative higher content of lactose compared to bovine milk. Table 1.2 Nutritional components in human milk, bovine milk, and formula Mature Human Milk Bovine Milk Infant Formula Total Protein (g/dL) 0.9-- 1.7 3.7 1.3- -1.5 Whey : Casein ratio 60 : 40 20 : 80 60 : 40 Total Fat (g/dL) 3.8 3.7 3 .5--3.7 Triglyceride (%total fat) 9 7 - -98 9 7 - 9 8 -Lactose (g/dL) 6.8--7 .3 4.8 6 .5--7.6 Vitamin C (mg/L) 5 0 - -100 11 7 0 - 84 Vitamin A (ug/L) 200 - 6 0 0 410 400 - 6 1 0 Vitamin E (mg/L) 2.5-- 4 . 0 7.0 17.4 - 1 8 . 4 (Picciano, 1998; Newton, 2004; U S D A National Nutrient Database, 2007) a Mead Johnson Enfamil®, low iron, ready-to-feed and Mead Johnson Enfamil L ip i l®, with iron, ready-to-feed, with A A and D H A . 1.1.3 Human Milk Protein Mature human milk protein consists of 30-40% casein and 60-70% whey protein, while bovine milk consists of 80% casein and 20% whey (Table 1.2). In comparison, infant formula has the same wheyxasein ratio as does human milk. Whey protein and casein distribution in human and bovine milk is shown in Table 1.3. Human milk is devoid of p"-lactoglobulin, which represents 20% of whey protein in bovine milk. Human milk is rich in 6 P-casein while bovine milk is rich in a-casein. The presence of milk protein, such as |3-lactoglobulin and a si-casein, in formulas that are derived from bovine milk is a risk factor for the onset of allergies in formula-fed infants (Malacarne et al., 2002). Lactoferrin and lysozyme are two protective factors that have immunomodulating, antiviral, and antibacterial activity (Worthington-Roberts, 1997; Picciano, 1998; Lindmark-Mansson and Akesson, 2000; Hamosh, 2001; Morrow et al., 2004; Newton, 2004). These antimicrobial factors are scarce in bovine milk. The presence of lactoferrin in human milk enhances iron absorption and reduces ferric iron in the gut of the breastfed infant (Newton, 2004) which can result in antioxidant protection (Lindmark-Mansson and Akesson, 2000). The content of K-casein in human milk is twice as much as that present in bovine milk. K-casein is a source of casein macropeptide (Wong et al., 2006), a strong promoting factor for the growth of protective bacteria Bifidobacterium bifidum, while inhibiting binding of pathogenic bacteria to gastric mucosa and epithelial cells of the respiratory tract (Hamosh, 2001). Table 1.3 Comparison of protein distribution of mature human milk and bovine milk Matu re H u m a n M i l k Bovine M i l k Whey Protein (g/dL) 0.8 0.6 P-lactoglobulin (%) Absent 1 8 - 2 0 a-lactalbumin (%) 3 0 - 4 5 5 3 - 5 4 Immunoglobulins (%) 1 5 - 2 0 1 0 - 1 2 Serum albumin (%) 5 - 9 6 - 8 Lactoferrin (%) 2 5 - 3 0 8 Lysozyme (%) 1 .5 -3 Trace Casein (g/dL) 0 . 4 - 0 . 6 2.5 a-casein (%) 11 - 12.5 4 8 - 4 8 . 5 P-casein (%) 6 3 - 7 7 3 6 - 3 8 K-casein (%) 2 2 - 2 5 1 3 - 1 4 Micel les size (nm) 64 182 Shown as percentage of whey protein or casein. (Hamosh, 2001; Malacarne et al., 2002) 7 Human milk protein provides the breastfed infant with essential amino acids and a unique amino acid composition that makes human milk distinct from bovine milk (Lawrence, 1994; Picciano, 1998; Newton, 2004). For example, human milk has a 1:1 methionine and cysteine ratio, a lower content of phenylalanine and tyrosine, and a high content of taurine (Lawrence, 1994; Newton, 2004). The low neonatal levels of specific enzymes to metabolize phenylalanine and tyrosine causes improper digestibility of bovine milk and formula (Lawrence, 1994). Taurine, which is almost absent in bovine milk but present in formula, has an important role as a neuromodulator in the brain and retina (Newton, 2004). Human milk is digested twice as fast as bovine milk or formula. The use of bovine milk protein to meet protein composition in formula causes difficulties on its digestibility for the infant (Newton, 2004), which is attributed to different factors that include protein composition, casein distribution, and the relative smaller micelle size of proteins. As a result digestion produces a finer, softer precipitate compared to bovine milk and formula (Malacarne et al., 2002). 1.1.4 H u m a n M i l k Carbohydrates The predominant carbohydrate of human milk is lactose, 3.8g/100mL. Glucose is the main energy source for the central nervous system developments and function (Lawrence, 1994; Newton, 2004). Complex oligosaccharides are also present and represent the third major component of carbohydrates in human milk. Approximately 80 oligosaccharides have been isolated from human milk while less than 20 have been described from bovine milk (Gopal and G i l l , 2000). These components appear to enhance host defenses by promoting the growth of nonpathogenic bacteria, especially Bifidobacterium bifidum (Newton, 2004). It has been also suggested that oligosaccharides in human milk inhibit binding of pathogens to epithelial surfaces (Hamosh, 2001) and prevent breastfed infants against diarrhea (Morrow et al., 2004). 1.1.5 H u m a n M i l k V i tamins The concentration of fat-soluble vitamins and carotenoids in human milk decreases with the stage of lactation (Macias and Schweigert, 2001). In mature human milk, vitamin A is comprised mainly of retinyl esters (Picciano, 2001) and more than 80% of total vitamin 8 E content is a-tocopherol (Picciano, 1998). Other isomers of tocopherol are present in smaller quantities. Vi tamin E content in infant formula is higher than in human milk, which protects formula from l ipid oxidation (Table 1.2). The concentration of water-soluble vitamins increases with lactation and reflects maternal diet and consumption levels (Lawrence, 1994). For example, vitamin C concentration in mature human milk is high compared to both bovine milk and infant formula (Picciano, 1998) (Table 1.2). More information on milk antioxidant vitamins is presented in Section 1.3. 1.1.6 H u m a n M i l k L ip ids M i l k lipids provide the breastfed infant with maximum intestinal absorption of fatty acids. About 50% of total calories from the milk consumed by the baby are derived from total milk lipids (Lawrence, 1994). Mature human milk fat is composed of 97-98% triglycerides, 0.7% phospholipids, and 0.5% cholesterol (Newton, 2004). The levels of some of the major fatty acids in human and bovine milk and formula are presented in Table 1.4. Human milk contains less saturated fatty acids from C4:0 to C10:0 and the percentage of unsaturated fatty acids is higher than in bovine milk. Fatty acids in relatively high concentration are oleic acid, palmitic acid, and linoleic acid ( LA ) with 33-46.4%, 20-30%, and 6-15%) of total fatty acids, respectively. The concentration of linolenic acid ( A L A ) in human milk ranges from 0.5-3.4% while maximum concentration of L A in bovine milk has been reported to be 1.8% of total fatty acids (Malacarne et al., 2002). The concentration of long-chain polyunsaturated fatty acids ( L C - P U F A ) in human milk is low. For example, content of arachidonic acid (AA) , eicosapentaenoic acid (EPA) , and docosahexaenoic acid (DHA) ranges from 0.5%, 0.1-0.2%), and 0.2-0.6% of total fatty acids, respectively. Bovine milk has very low levels of these L C - P U F A and most infant formulas are enriched with A A , E P A , and D H A in concentrations of 0.5, 0.1, and 0.35%, respectively. Table 1.4 exemplifies one brand of formula and L C - P U F A s have lower concentrations than those of human milk. In the production of infant formulas, special attention is given to minimize E P A , since it may antagonize the co-6 A A metabolism and thus interfere with infant growth (Simopoulos et al., 1999). Fat is the most variable constituent of human milk and can be influenced by different factors such as duration of gestation, stage of lactation, parity, nursing, and maternal energy 9 status; however, maternal diet is the main factor (Picciano, 200,1). Fat content in human milk increases after 3 to 5 minutes of nursing the baby, this mi lk is known as hindmilk (Lawrence, 1994; Picciano, 2001), while foremilk is the milk eaten during the first minutes of nursing (Turoli et al., 2004). A shorter gestation period can result in an increased secretion of L C -P U F A in colostrum, and this first milk has a relative higher content of phospholipids and cholesterol than transitional and mature milk (Picciano, 2001). Table 1.4 Fatty acid composition of mature human milk, bovine milk, and infant formula. Shown as % of total fatty acids Mature Human Milk Bovine Milk Infant Formula3 Total Fat (g/dL) 3.8 3.7 3.7 Total fatty acids 88 88 (% total fat) Butyric (C4:0) 0.1 1 .4-3 .0 0.0 Caproic (C6:0) 0.2 1 .6-2 .2 0.01 Caprylic (C8:0) 0 . 1 - 0 . 3 1 .3-1 .8 0.06 Capric (C10:0) 1.0-2.1 3 . 0 - 3 . 6 0.04 Laurie (C12:0) 3 . 0 - 7 . 2 3 . 0 - 4 . 0 0.33 Myrist ic (C14:0) 5 . 0 -10 .0 13 .0 -14 .0 0.15 Palmitic (C16:0) 2 0 . 0 - 3 0 . 0 2 9 . 5 - 4 3 . 0 0.8 Stearic (C 18:0) 6 . 0 - 9 . 0 5 .7 -13 .7 0.15 Palmitoleic (C16: l ) 3.5 1.7 1.3 Oleic (C18: l ) 33 .0 -46 .4 16.7-27.1 1.3 L A (C18:2 co-6) 6 . 0 - 1 5 . 0 1 .6-3 .0 0.6 A A (C20:4 co-6) 0.5 0.0 - trace 0.02 A L A (C18:3 co-3) 0 . 6 - 3 . 4 0 . 5 - 1 . 8 0.06 E P A (C22:5 co-3) 0 . 1 - 0 . 2 0.0 0.0 D H A (C22:6 co-3) 0 . 2 - 0 . 6 0.0 - trace 0.01 (Boersma et al., 1991; Francois et al., 2001; Picciano, 2001; Malacarne et al., 2002 ; U S D A National Nutrient Database, 2007) a Mead Johnson Enfamil L ip i l ®, with iron, ready-to-feed, with A A and D H A . 10 Maternal dietary fatty acid intake during pregnancy and lactation greatly influences the fatty acid composition of human milk (Innis, 2003; Si lva et al., 2005; Helland et al., 2006; Hadders-Algra, 2007). For example, a high maternal intake of L A during pregnancy was observed to decrease the D H A concentration in human milk compared to women with a low intake of L A (A l et al., 1995). Higher concentrations of saturated and monounsaturated fats in human mi lk can reflect a high carbohydrate and low fat diet, which results in endogenous synthesis of medium chain fatty acids (Picciano, 2001; Agostoni, 2005). 1.2 Fatty Ac ids and Neonatal Heal th As mentioned in Section 1.1.6, both saturated and unsaturated human milk fatty acids provide almost 50% of total energy to the breastfed infant. Polyunsaturated fatty acids (PUFA) are classified on the basis of a chemical structure that includes omega-3 (co-3) or omega-6 (co-6) fatty acids. These P U F A have a double bond at carbon three or carbon six from the methyl carbon, respectively (Figure 1.1). Important fatty acids in infant nutrition include a-linolenic acid ( A L A , C18:3 co-3), eicosapentaenoic acid (EPA, C20:5 co-3) and docosahexaenoic acid ( D H A , C22:6 co-3) from the omega-3 group, and linoleic acid (LA , C18:2 co-6) and arachidonic acid ( A A , C20:4 co-6) from the omega-6 group (Innis, 2003; Ettinger, 2004). Figure 1.1 Chemical structure of the essential fatty acids omega-3 a-linolenic acid (A) and omega-6 linoleic acid (B) 11 A A , E P A and D H A are formed from precursor essential fatty acids as shown in Figure 1.2 (Innis, 2003; Ettinger, 2004; Balk et al., 2006). Essential fatty acids include L A (omega-6) and A L A (omega-3); however, there is some controversy regarding whether the rate of conversion is adequate to meet neonatal needs. It has been suggested that the A L A content of human milk may not be sufficient to meet the requirements for D H A synthesis in the absence o f adequate preformed D H A (Innis et al., 2001). These findings suggest that newborn infants are capable of the metabolic fatty acid conversion processes, but synthesis of long-chain polyunsaturated fatty acids ( L C - P U F A ) such as D H A , is limited. Diet Omega-6 fatty acids Omega-3 fatty acids C18:2 W-6 C18:3 co-3 | •*— A6 desaturase —*• j C18:3 00-6 C18:4 C O - 3 | -*— elongase — • j C20:3 co-6 C20:4 co-3 | «— A5 desaturase —»* j C20:4 C O - 6 C20:5 C O - 3 | — elongase —> C22:4 C O - 6 C22:5 co-3 | •«— A4 desaturase — > j C22:5 C O - 6 C22:6 C O - 3 Figure 1.2 Scheme of omega-6 and omega-3 fatty acid desaturation and elongation 12 L C - P U F A s are important for fetal and infant growth and visual and neural development. Two critical L C - P U F A are A A and D H A . A A is found in cell membrane phospholipids and is important in secondary messenger and cell signaling pathways, in cell division, and as a precursor for eicosanoid synthesis. A A is also known to reverse growth retardation associated with deficiency of essential fatty acids (Ettinger, 2004; Newton, 2004; Hadders-Algra et al., 2007). On the other hand, D H A is involved in visual acuity, neural development and function, neurotransmitter metabolism, and behavior (Newton, 2004; Hadders-Algra et al., 2007; Watkins et al., 2007). Visual acuity was found to be significantly better in breastfed than in formula fed infants (Kurlak and Stephenson, 1999). E P A and A A are competitors in their metabolic pathways; for example E P A reduces inflammation and vasoconstriction, while A A can stimulate those responses (Innis, 2003). E P A is a natural constituent in human milk; however high infant E P A intake has been associated with low growth in preterm infants fed formula supplemented with h igh-EPA fish oils (Carlson et al., 1992). Although P U F A are subject to oxidative decomposition due the presence of unique structural double bonds separated by methylene groups (Banks, 1997; Schneider, 2005), P U F A have been studied to also evaluate a potential antioxidant effect in biological enriched systems (Vis io l i et al., 1998; Wander, 2001; Barbosa et al., 2003; Mor i , 2004). Studies on dispersions of lipids in aqueous solutions to simulate biological systems reported that omega-3 L C - P U F A s have relative greater oxidative stability than omega-6 L C - P U F A s (Wander, 2001). Using conjugated dienes (CD) as a measure of l ipid peroxidation after radical-induced oxidation, it was found that the loss of D H A was relatively slower than E P A , A A , A L A , respectively; while L A was the most susceptible to oxidization (Visiol i et al., 1998). Potential antioxidant activity pathways of omega-3 fatty acids have been proposed. A decreased in plasma oxidative stress, suggested that omega-3 L C - P U F A may act as sacrificial free radical scavenger that protect patients with ulcerative colitis against the overall effects of oxidative stress (Barbosa et al., 2003). In addition, Mor i (2004) suggested that presence of omega-3 P U F A in membrane lipids and lipoproteins may also inhibit the pro-oxidant enzyme phospholipase A 2 and stimulate activity of antioxidant enzymes. The effect on human milk oxidative stability due to different levels of unsaturation and the presence of individual fatty acids has not been studied. 13 1.3 Antioxidant System in Human Milk and Neonatal Health Recently, it has been considered that one of the most important aspects of human milk is the role to provide an antioxidant defense system to breastfed babies. Neonatal oxidative status reflects both the maternal nutritional status and the physicochemical aspects of individual soluble milk constituents. The antioxidant system of human milk comprises numerous bioactive components found to have antioxidant activity, such as vitamins A , E, and C , peptides like lactoferrin, lysozyme and casein, and the enzymes catalase (CAT) , superoxide dismutase (SOD), and glutathione peroxidase (GPx) which are important for detoxifying R O S in the breastfed infant (L 'Abbe and Friel , 2000; Lindmark-Mansson and Akesson, 2000; Hamosh, 2001; Friel et al., 2002; Shoji et al., 2004; Kasapovic et al., 2005; Kitts, 2005; Shoji et al., 2005, Wong et al., 2006). Friel et al. (2002) studied the protection against l ipid damage of antioxidant enzymes present in human milk, such as C A T , SOD, and GPx. A n increased protection against oxidative damage and l ipid oxidation was found when the antioxidant enzymes were added to infant formula. Human milk was found to suppress oxidative stress and oxidative D N A damage in newborn infants after 14 days of age, but a similar result was not found in infant formula (Shoji et al., 2004). These results suggest that human milk contains a unique defense mechanism which is not as available in commercial infant formulas and in bovine milk. A study reported human mi lk to be effective at protecting against hydrogen peroxide-induced oxidative damage in an intestinal epithelial cell line (IEC-6) (Shoji et al., 2005). This finding suggests an antioxidant mechanism in the gastrointestinal tract during infancy which protects neonates against diseases such as necrotizing enterocolitis. 1.3.1 Total Antioxidant Capacity of Human Milk The total antioxidant capacity (TAC) of human milk represents a measure of the total antioxidant activity, and thus milk oxidative stability, including all possible interactions among different compounds that can reduce oxidative reactions. T A C has been measured using of different methods (VanderJagt et al., 2001; Alberti-Fidanza et al., 2002; Hanna et al., 2004). M i l k T A C was positively correlated to maternal total antioxidant status of Nigerian women (r=0.62 p< 0.001) (VanderJagt et al., 2001). A former study conducted in Italy in 2002, determined the T A C of colostrum (within 72 hours postpartum), transitional 14 milk (at 8 days postpartum) and mature milk (at 20 days postpartum). It was suggested that T A C value of mature mi lk is lower than those values of colostrum and transitional milk. In addition, it was found that two partcipants who had smoked during pregnancy exhibited lower milk T A C value than those of participants who did not smoke (Alberti-Fidanza et al., 2002). The effect of storage on human milk T A C has also been evaluated by the Trolox Equivalent Antioxidant Capacity ( T E A C ) assay (Hanna et al., 2004). Preterm and term milk samples were collected within 24hr postpartum and T A C values were compared to those of commercial infant formulas. A decrease on T A C was suggested with refrigerator and freezer storage in both human milk and infant formula. The T A C of human milk was higher than that of formula; however, this result is contradictory with the findings from Alberti-Fidanza et al. (2002), that suggested a higher T A C in formulas compared to human milk. 1.3.2 V i tam in C and Neonatal Heal th Vi tamin C is the L-enantiomer of ascorbic acid and exists as two inter-convertible compounds: L-ascorbic acid and L-dehydroascorbic acid (Figure 1.3). The activity of vitamin C is multifunctional being an excellent reducing agent, metal sequester, and oxygen scavenger that traps both singlet oxygen and superoxide anion (Kitts, 1997). Figure 1.3 Strucure of L- ascorbic acid and L-dehydroascorbic acid Ortega et al. (1998) suggested that the levels of vitamin C in human milk are affected by dietary vitamin intake, and this may vary depending on the season, availability of vitamin C-containing food sources, and the level of consumption of fruits and vegetables. Another study concluded that the regular consumption of fresh orange juice at 3 or 5 times a week 15 (300 to 500mg ascorbic acid per serving) effectively increased human milk ascorbic acid 2-fold (Daneel-Otterbech et al., 2005). During pregnancy, vitamin C transport occurs across the placenta by active transport, resulting in a higher concentration in placental tissue or cord blood than in either fetal or maternal blood (Ladipo, 2000; Baydas et al., 2002). A higher concentration of vitamin C in fetal plasma has been reported at birth with a significant decline occurring postnatally (Zoeren-Grobben et al., 1994; Baydas et al., 2002). High concentrations of vitamin C at birth are of questionable benefit because vitamin C can act as a pro-oxidant by inhibiting ferroxidase activity of ceruloplasmin. On the other hand, vitamin C may protect against oxidative damage when the newborn goes from a relative hypoxic to a hyperoxic environment at birth (Zoeren-Grobben et al., 1994). 1.3.3 V i tam in A and Neonatal Heal th Vi tamin A is a fat-soluble vitamin that belongs to the family of chemical compounds known as retinoids; it is derived from isoprene and has an alcohol functional group (Figure 1.4). Beta-carotene (P-carotene) is the most known carotenoid and its vitamin A activity occurs upon conversion into retinol (Gallagher, 2004). A n important function of vitamin A involves the production of rhodopsin, a light sensitive pigment in the eye; the need for normal epithelial cell differentiation and function of the skin, mucous membranes, and blood vessel walls; it is essential to maintaining the immune system and prevent infections. Vitamin A also has antioxidant activity that can be characterized as a peroxyl radical scavenger, which is effective in reducing l ipid peroxidation (Tesoriere et al., 1996). Some similarity with vitamin E, as a chain breaking antioxidant, has been suggested, but it does not protect from superoxide or hydroxyl radicals (Bohles, 1997). P-carotene is a more effective scavenger of singlet oxygen than vitamin E, but exhibits antioxidant activity only at low oxygen concentration (Banks, 1997; Ternay and Sorokin, 1997). Figure 1.4 Structure of all-fram'-retinol 16 Maternal dietary intake of vitamin A generally reflects the concentration of vitamin A in human mi lk (Olafsdottir et al., 2001); however, some contradictory evidence exists regarding this finding (Oostenbrug et al., 1998; Meneses and Trugo, 2005). Most vitamin A from the diet is stored in the liver, and a requirement by a particular part of the body wi l l result in mobilization for use in various target cells and tissues (Gallagher, 2004). Since retinol storage in the fetal liver increases during late pregnancy, the fetus is protected against vitamin A deficiency (Baydas et al., 2002). Fetal blood vitamin A concentrations are maintained until the entire liver depot is utilized; however, premature or low birth weight infants are at a high risk to developing deficiency (Bohles, 1997; Baydas et al., 2002). Vitamin A is also very important for fetal lung development and maturation. A t late gestation, fetal maturation occurs with rapid depletion in retinol stores (Bohles, 1997). 1.3.4 V i t am in £ and Neonatal Heal th Vi tamin E exists as eight different forms or isomers; four tocopherols and four tocotrienols. A l l isomers have a chromanol ring, with a hydroxyl group which can donate a hydrogen atom to reduce free radicals and a phytyl tail with a hydrophobic side chain that allows for penetration into biological membranes. There is an alpha (a-), beta (P-), gamma (y-), and delta (5-) form of both the tocopherols and tocotrienols, which is determined by the number of methyl groups on the chromanol ring (Traber, 1999). The difference between tocopherols and tocotrienols is the unsaturation of the phytyl tail in tocotrienols. The unsaturated tail w i l l facilitate passage through the membrane bi-layer in a more efficient manner, thus resulting in a faster uptake compared to the tocopherol. Maximal uptake levels of tocotrienols are however lower than tocopherols (Tsuzuki et al., 2007). Figure 1.5 shows the structure of a-tocopherol. Each vitamin E isomer has a specific biological activity which wi l l depend on the binding specificity to the tocopherol binding protein in the liver (Schneider, 2005). The only form of vitamin E that is actively maintained in the human body is a-tocopherol and therefore, it is found in the largest concentrations in both blood and tissues (Traber, 1999). a-tocopherol is biologically and chemically the most active form of vitamin E; largely because of the preference for the transport mechanism present in the liver (Schneider, 2005). Moreover, the substitution pattern of methyl groups on the chromanol ring in a-tocopherol is 17 very active by virtue of its affinity, and facilitates the transfer of a hydrogen atom to a peroxyl radical ( R O O ) (Figure 1.6). In l ipid systems, a-tocopherol acts as a radical scavenger and terminates the propagation of radical chain reactions by reacting with peroxyl radicals and generating unreactive phenoxyl radicals and hydroperoxide products (ROOH) (Kitts, 1997; Schneider, 2005). Chromanol ring Phytyl tail Figure 1.5 Structure of a-tocopherol a-tocopherol + Peroxyl radical • Tocopheroxyl + Hydroperoxide ROO- radical ROOH Figure 1.6 Scheme of the antioxidant activity of a-tocopherol against peroxyl radicals A higher maternal intake of vitamin E can lead to a higher level of the same vitamin in maternal circulation and in human milk (Ortega et al., 1999). Vi tamin E deficiency in newborn infants has been associated with hemolytic anemia, bilirruninemia, intracranial hemorrhage, and thrombocytosis (Bohles, 1997; Sokol, 1988). In additition, pre-term infants are predisposed to a deficiency since vitamin E accummulates in the fetus mostly during the third trimester of gestation (Baydas et al., 2002). The antioxidant activity of vitamin E is 18 important during gestation, since there is an increase in maternal plasma lipid levels and it has been suggested that an accompanied increase in maternal plasma vitamin E levels protects against in vivo l ipid peroxidation (Oostenbrug et al., 1998). In addition, vitamin E becomes important when maternal intakes of P U F A are high (Ortega et al., 1999; Olafsdottir etal . , 2001). 1.4 Oxidat ive Stress and Neonatal Heal th Oxidative stress has an important role in the etiology of many diseases such as cancer, coronary heart disease, and atherosclerosis in adults; in pregnancy related disorders such as pre-eclampsia and intrauterine growth restriction (Pressman et al., 2003; Borna et al, 2005). Oxidative stress occurs when the antioxidant system cannot neutralize, remove, or transform R O S to less toxic species (Salem and Baskin, 1997). A t birth, the transition of the neonate from the relative hypoxic utero state to the hyperoxic environment results in elevated R O S production due to the exposure to aerobic oxygen and the activation of the oxidative metabolism (Shoji et al., 2005). Since vitamins are transported to the fetus predominantly in the third trimester of gestation and expression of antioxidant enzymes is relatively low at birth, newborn infants are at a higher risk of oxidative stress injury, especially those that are born prematurely (Bohles, 1997; Oostenbrug et al., 1998; Baydas et al., 2002). A higher level of D N A oxidation, measured by 8-OHdG was found to be present in premature infants, and corresponded to lighter body weights (Matsubasa et al., 2002). Excessive exposure of the neonate to R O S resulting in enhanced oxidative stress can cause intraventricular hemorrhage ( IVH), retinopathy of prematurity (ROP), necrotizing enterocolitis (NEC) , acute tubular necrosis, and chronic lung disease (CLD) (Matsubasa et al., 2002; Shoji et al., 2004; Shoji et al., 2005). In addition, neonatal antioxidant defense system against R O S is not completely developed, thus the infant relies on human milk to provide with the antioxidant protection (L 'Abbe and Friel , 2000; Lindmark-Mansson and Akesson, 2000; Hamosh, 2001; Shoji et al., 2004; Kasapovic et al. , 2005; Shoji et al., 2005), as discussed in Section 1.3. However, l ipid peroxidation in human milk may cause the ingestion of oxidated compounds that wi l l affect the oxidative balance of the breastfed baby (Miranda et al., 2004; Turoli et al., 2004). 19 1.5 Oxidative Stability of Human Milk: The Role of ROS and Lipid Peroxidation Reactive oxygen species (ROS) are free radicals and peroxides which are generated from a number of both endogenous and exogenous sources. Primary R O S include superoxide radical (O2') , hydroxyl radical (OH'), and hydrogen peroxide (H2O2) and peroxyl radical (ROO') . A n endogenous source of R O S is the respiratory chain. During respiration, oxygen is reduced to water; however, this process also results in the formation of R O S , which function as aggressive electron donors (Figure 1.7). Excessive R O S production wi l l increase l ipid, protein, and D N A oxidation (Hall iwell and Gutteridge, 1999) in cell membranes, body fluids, or food systems such as human milk (Bohles, 1997). H2O « O2 \ Figure 1.7 Scheme of production of reactive oxygen species (ROS) in the cell H 0 2 H2O2 H2O 02-" H2O2 H O In all systems (cell membranes, body fluids, or food systems), a primary target for R O S interaction is P U F A (Schneider, 2005). The initiation reaction is characterized by the abstraction of a hydrogen atom from a methylene carbon that is located in between two double bonds of the P U F A (RH) and yields a carbon-centered radical (R») (Figure 1.8 (1)). Once the carbon radical is formed, it wi l l immediately react with molecular oxygen to form an unstable fatty acid peroxyl radical (ROO») (Figure 1.8 (2)). This reaction is relatively fast and is the start of a chain reaction, where the peroxyl radical reacts with another fatty acid (by hydrogen abstraction) to form more peroxyl radicals (Figure 1.8 (3)). This chain reaction continues to termination, which can occur by the donation of a hydrogen atom to the peroxyl radical from chain breaking antioxidants (TH) and by the reaction of the peroxyl radical and an antioxidant radical (Figure 1.8 (4)). 20 Initiation: RH + X — * R- + XH (1) Tennination: ROO- + TH —> ROOH + T- (4) ROO- + T- — * ROOT Figure 1.8 Scheme of reactions in l ipid peroxidation Oxidative stability of human milk has been studied at different conditions of storage. A decrease in milk T A C was found at 4°C and at -20°C (Hanna et al., 2004; Miranda et al., 2004; Turoli et al., 2004). In addition, the oxidative status of human milk has also been reported to be higher than that of infant formulas and the increased concentration of lipid peroxidation products was attributed to the presence of L C - P U F A and active bile-salt stimulated lipase (BSSL) in human milk (Turoli et al., 2004). However, high levels of lipid peroxidation products did not correspond to lower T A C of human milk, suggesting that P U F A are not strictly related to milk T A C (Turoli et al., 2004). 21 1.6 Measures of Oxidative Status Since R O S are not specific to one particular biomolecule, oxidative stress wi l l result in damage of D N A , protein, and lipids. There are specific methods which can be used to assess quantitatively or qualitatively the extent of oxidation in various macromolecules. 1.6.1 DNA oxidation A primary result of oxidative damage to D N A is the generation of 8-hydroxy-2'-deoxyguanosine (8-OHdG) and the free base 8-hydroxyguanine (Hwang and K i m , 2007). Studies conducted with urine and D N A extracted from cells, such as circulating leukocytes, have measured the extent of oxidative damage repair or the balance between damage and repair (endogenous 8 -OHdG level) (Therond et al., 2000). Levels of 8 -OHdG are determined by E L I S A , based on monoclonal antibody or by using a solid phase extraction and liquid chromatography with electrochemical detection (Therond et al., 2000; Matsubasa et al., 2002). The former method has been used to study the oxidative status of premature infants during intensive care. For example, rather than finding that administered oxygen to those infants was the only cause for excessive oxidative damage, it was concluded that the levels of urinary 8-OHdG corresponded to the degree of prematurity in premature infants (Matsubasa et al., 2002). 1.6.2 Protein oxidation R O S also lead to oxidation of amino acid residue side chains, formation of protein-protein crosslinking or fragmentation of the protein backbone. Among amino acid residues in proteins susceptible to oxidation, lysine, arginine, proline and threonine yield carbonyl derivatives (Therond et al., 2000; Hwang and K i m , 2007). Carbonylated proteins in plasma are biomarkers of oxidative stress; however, this method is time consuming and loss of acid soluble proteins can occur during extraction (Therond et al., 2000). Carbonylated proteins can be measured by colorimetry based on the reaction of the carbonyl group with dinitrophenylhydrazine (Hwang and K i m , 2007), with immunologic methods on tissue or isolated protein samples, using an E L I S A assay, by gas chromatography/mass spectroscopy, and by liquid chromatography after isolation, purification and hydrolysis of the protein under study (Davies et al., 1999). 22 During exercise, oxidative stress is known to induce increased production of stress proteins or heat-shock proteins (HSPs) in blood cells, such as lymphocytes. HSP60 and HSP70 are two specific HSPs that have been used as biomarkers for assessing oxidative stress both prior and fol lowing exercise (Jackson et al., 2004). 1.6.3 L i p i d oxidation Thiobarbituric acid reactive substances ( T B A R S ) is the most frequently employed method to measure l ipid peroxidation. The substrate is malondialdehyde ( M D A ) , a major breakdown product of l ipid peroxides (Morris et al., 1998). M D A derives from b-cleavage of endocyclization of arachidonic acid hydroperoxide and reacts with thiobarbituric acid to give a pink and fluorescent chromogen that is measured by colorimetry or fluorimetry. This method lacks specificity for M D A since all compounds with reducing activity wi l l react with thiobarbituric acid (Therond et al., 2000; Hwang and K i m , 2007). Conjugated dienes (CD) are formed during the early steps of l ipid peroxidation, characterized by a double-single-double bond. These structures absorb ultraviolet light in the range 230 - 235nm, which allows measurement at this absorbance. However, many different substances absorb ultraviolet light (e.g., heme proteins, purines, pyrimidines, carbonyl compounds formed upon l ipid peroxidation), and so this measurement cannot be performed directly on serum or plasma unless products are extracted from lipids into organic solvents (Therond et al., 2000). Estimation of low density lipoprotein ( L D L ) oxidation in vivo and in vitro has been based on the C D assay (Hu and Kitts, 2001; Prior, 2004). L ip id hydroperoxides, measured as cholesteryl ester hydroperoxides, are the most abundant plasma l ipid hydroperoxides generated in vivo and formed in vitro (Therond et al., 2000). Lipids are extracted in solvent and hydroperoxides are detected by liquid chromatography with chemiluminescence detection, or by using the ferrous oxidation in xylenol orange (FOX-2 assay) (Nourooz-Zadeh et al., 1996; Morris et al., 1998). Isoprostanes are a family of eicosanoids of non-enzymatic origin, produced by the free radical oxidation of tissue phospholipids, specifically arachidonic acid ( A A ) (Mastaloudis et al., 2001). F2-isoprostanes are chemically stable end-products of l ipid peroxidation and important biomarkers of oxidative stress which can be detected in all tissues and body fluids by gas chromatography/mass spectrometry and immunoassays (Hwang and K i m , 2007). '23 1.7 Measures of Total Antioxidant Capacity Due to the complexity of the biological antioxidant system, and the possible interaction among different antioxidants, measurement of a single antioxidant or measurement of each antioxidant separately wi l l be less representative than the overall antioxidant status. It is also possible that synergy between a number of antioxidants working in different ways results in an enhanced activity (Kitts, 1997; Rietjens et al., 2002; El is ia et al., 2007). Total antioxidant capacity (TAC) is a measure of the total antioxidant activity of the different compounds present in a sample or system. For purposes of this thesis research, emphasis has been given to the Oxygen Radical Absorbance Capacity ( O R A C ) assay. 1.7.1 DPPH Radical Scavenging Activity In this assay, the antioxidant activity of a compound is measured by the reduction of a stable free radical 2,2-diphenyl-l-picrylhydrazy ( D P P H ) (Bondet et al., 1997; Kitts et al., 2000; Schlesier et al., 2002; Mil iauskas et al., 2003). In its radical form, D P P H - shows an absorbance at 515nm which disappears upon reduction. The antioxidant activity is calculated by determining the decrease in absorbance as percentage of inhibition (Kitts et al., 2000; Schlesier et al., 2002). 1.7.2 Ferric Reducing Antioxidant Power (FRAP) This assay measures the reducing power of the compound in respect to its antioxidant function. It measures the reduction of ferric (Fe 3 + ) to ferrous (Fe 2 + ) complex at low pH, which forms a colored ferrous-tripyridyltriazine complex (Yeum et al., 2004). F R A P is a quick and simple method; however, neither F e 3 + nor F e 2 + is able to directly cause oxidative damage to lipids, proteins or nucleic acids (Prior and Cao, 1999). 1.7.3 Trolox Equivalent Antioxidant Capacity (TEAC) The Trolox Equivalent Antioxidant Capacity ( T E A C ) assay measures the affinity of a compound in reducing the radical cation of 2,2'-azinobis(3-ethylbenzothiazoline 6-sulfonate) (ABTS) (Prior and Cao, 1999). The procedure generally measures the reducing power of the compound and compares it in respect to its antioxidant function. The A B T S radical is a stable and pre-formed radical, but it does not act as pro-oxidant (Yeum et al., 2004). 24 Disadvantages of this assay include that A B T S is not found in mammalian biology and thus represents a non-physiological radical; a compound can only reduce A B T S i f it has a redox potential lower than that of A B T S , and the T E A C values may not be the same for slow reactions taking a long time to reach the endpoint (Phipps et al., 2007) 1.7.4 Total Radical Trapping Parameter (TRAP) The Total Radical Trapping Parameter (TRAP) assay measures the oxygen consumed during the reaction between peroxyl radicals and the test sample. This assay uses R-phycoerythrin (R-PE) or 2',7'-dichlorodihydrofluorescein (DCFH) as the oxidizable substrate (fluorescent probe) and either 2,2'-azobis(2-amidinopropane) dihydrochloride ( A A P H ) or A B T S as initiators of free radical generation (Yeum et al., 2004: Huang et al., 2005). Calculations of antioxidant capacity with this assay are based only upon the observed length of the induction period (lag phase) (Cao and Prior, 2002; Huang et al., 2005); however, lag time-based measurements of antioxidant capacity are found to overestimate the antioxidant capacity of weaker antioxidants (Prior, 2004). In addition, the oxygen electrode wi l l not maintain its stability over the period of time required (Cao and Prior, 2002). 1.7.5 Oxygen Radical Absorbance Capacity (ORAC) This assay is based on the capacity of antioxidants in a sample to directly quench induced peroxyl radicals (pro-oxidants) (Prior and Cao, 1999). Induced oxidation is initiated by thermal decomposition at 37°C of A A P H , which produces carbon-centered radicals that then react with oxygen yielding to the corresponding peroxyl radicals (Yeum et al., 2004; Kitts and Hu, 2005). The assay measures the oxidative degradation of a fluorescent probe. The protein B-phycoercythrin (B-PE) was initially used as the fluorescent probe; however, its variable reactivity and lack of photostability determined the need for an improved method by using fluorescein (FL) as the probe (Ou et al., 2001). The assay using F L is known as O R A C F L - The fluorescent intensity of F L decreases with oxidization, but the oxidative decay is inhibited in the presence of antioxidant activity in the sample under analysis. The O R A C value of a sample is calculated by using a standard curve (Trolox Calibrator) derived from Trolox, a water-soluble vitamin E analogue. The O R A C value is usually reported as (umol Trolox/unit of sample) or "Trolox equivalents" (Huang et al., 2005). 25 The O R A C assay simulates a biological system and utilizes a biological relevant radical source. Reaction is driven to completion, to where all non-protein antioxidants (vitamins, uric acid, and bilirubin) and most of the protein (including albumin) in the sample are oxidized by the peroxyl radicals (Cao et al., 1993). Decomposition of A A P H and production of peroxyl radicals undergo at a rate that is primarily determined by temperature and the presence of the test sample, thus the O R A C assay has high specificity (Prior and Cao, 1999). The use of F L as the flourescent probe allows this assay to be inexpensive and very photostable, since F L does not interact with antioxidants or other reactive compounds (Ou et al., 2001). The ORACR-PE assay has been used to evaluate the T A C of human milk at different stages of lactation (Alberti-Fidanza et al., 2002); however, evidence regarding the standardization and validation of the assay to measure the total antioxidant capacity for evaluating human mi lk oxidative stability has not been reported. 26 C H A P T E R 2: S T A N D A R D I Z A T I O N A N D V A L I D A T I O N O F T H E O R A C F L A S S A Y F O R T H E M E A S U R E M E N T O F O X I D A T I V E S T A B I L I T Y O F H U M A N M I L K 2.1 INTRODUCTION Total antioxidant capacity (TAC) of human milk is a measurement of the total antioxidant activity of bioactive components that are present in mi lk and provide oxidative stability. M i l k T A C has been measured by different methods (VanderJagt et al., 2001; Alberti-Fidanza et al., 2002; Hanna et al., 2004). The affinity of human milk at reducing the radical cation A B T S has been measured by the T E A C assay to evaluate the antioxidant content of milk from Afr ican women (VanderJagt et al., 2001) and to evaluate the effect of storage on milk antioxidant activity (Hanna et al., 2004). However, the T E A C procedure is limited by the fact that A B T S represents a non-physiological radical and a compound can only reduce A B T S i f it has a redox potential lower than that of A B T S (Phipps et a l , 2007). The Oxygen Radical Absorbance Capacity ( O R A C ) assay is a widely used procedure that measures the total antioxidant capacity of biological samples, supplements, and food samples (Prior et al., 2003; Davalos et al., 2004; W u et al., 2004; Fernandez-Pachon, et al., 2005). The O R A C F L assay has been proposed as a more advantageous method due to its sensitivity, the photostability of the fluorescent probe fluorescein (FL) , and the specificity for antioxidant activity against peroxyl radicals (Ou et al., 2001). The O R A C assay using the fluorescent probe B-phycoercythrin (B-PE), has been previously used to measure the antioxidant capacity of colostrum, transitional and mature human milk in one study from Italian women (Alberti-Fidanza et al., 2002); however, there is no evidence that this assay was standardized or validated for use in human milk. The work described in this chapter was designed and conducted to standardize and validate the improved O R A C F L assay for assessing the oxidative stability of human milk. The experiments were designed to standardize the protocol and determine the limits of linearity and precision of the standard calibrator and the accuracy of the method. Recovery experiments of known antioxidant added to milk and determining appropriate milk dilution for measuring O R A C F L in the human milk matrix were performed. 27 2.2 MATERIALS AND METHODS 2.2.1 Human Milk Samples For the standardization of the O R A C F L assay, the B C Women's M i l k Bank provided a sample of human mi lk constituted from pooled milk collected from different healthy unknown women. Pooled milk was stored at -80°C. Frozen milk was thawed and aliquots were placed in sterile tubes and transported on ice to the Food Science Laboratory and immediately frozen at -20°C. Whole milk was used in all analyses. 2.2.2 Materials and Equipment 75mM Phosphate buffer (PB), pH=7.4, was made from the dilution 1:9 (v/v) of a pre-made mix 61:39 (v/v) of 0.75M K 2 H P 0 4 and 0.75M N a H 2 P 0 4 . P B was stored in refrigerator. Trolox Calibrator 20uM Trolox was made in 75mM P B and stored at -20°C until used. Fluorescein sodium salt (FSS) stock solution was made in 75mM P B and stored at -20°C until used. Further dilutions were needed to achieve 200nM FSS. The peroxyl radical initiator 2,2'-azobis(2-amidinoproprane) di-hydrochloride ( A A P H ) was made in 75mM P B immediately before use. A l l chemicals were purchased from Sigma-Aldrich. 96-well black assay plates were purchased from BioRad (Cat.No.353241). Fluorescence was taken in the Fluoroskan Ascent F L (Labsystems) at excitation 485nm and emission 527nm. 2.2.3 Standardization and Validation of the O R A C F L Assay 2.2.3a The O R A C F L Assay The improved O R A C F L assay was standardized and validated for human milk based on the methodology described by Ou et al. (2001) and by Kitts and H u (2005). The assay was run several times (i.e. 20 times) in order to become familiarized and trained before the actual standardization and validation experiments were performed. Table 2.1 shows the specific amounts of each chemical used in the assay. Six different concentrations of Trolox standard (0.0, 0.5, 1.0, 2.0, 3.0 and 4.0uM), a water-soluble vitamin E analogue, were used to make the Trolox Calibrator. To start with the O R A C F L assay, solutions of different chemicals were prepared as specified in Section 2.2.2. As presented in Table 2.1, specific volumes of Trolox standard, P B , and FSS were added to a 96-well microplate for each concentration of Trolox Calibrator, to a final volume of 160uL. The microplate was shaken 28 for lOsec (600rpm), and incubated at 37°C. The radical inducer A A P H was prepared just before being used. When a temperature of 37°C was reached, a volume of 40uL of A A P H was added to all wells of the plate, except for the control and the blank. A control sample consisted of 140uL P B and 60uL F S S , while the blank contained only 200uL P B . The final volume of the assay was 200uL. The microplate was shaken for 10 sec (600rpm), and fluorescence was read every minute for 60 minutes at excitation 485nm and emission 527nm. Fluorescence values were obtained at the end of the analysis. Area under the curve values (AUC) were calculated with the integration from time 0 to 60 min following the formula A U C = 0.5 + S A i / A l + 0.5(A60/A0), where A is the fluorescence value, ti=0 min and tj=60 min. Data integration and calculations were performed according to Davalos et al. (2004). Table 2.1 Specific chemicals needed to make Trolox Calibrator in O R A C F L assay 20uM Trolox 75nM PB 200nM F S S 6 0 n i M A A P H Trolox (ML) (UL) (UL) (ML) concentration (MM) Blank 0 200 0 0 Control 0 140 60 0 Trolox 0 100 60 40 0.0 Calibrator 5 95 60 40 0.5 10 90 60 40 1.0 20 80 60 40 2.0 30 70 60 40 3.0 40 60 60 40 4.0 P B = Phosphate buffer; FSS = Fluorescein sodium salt A A P H = 2 , 2 ' - azobis(2-amidinoproprane) di-hydrochloride M i l k samples and Trolox Calibrator net A U C values were obtained by subtracting the A U C value of the blank from that of milk sample or Trolox Calibrator. Sample and Trolox Calibrator net A U C values were plotted versus concentration, [umol Trolox/mL] for Trolox Calibrator and [mL mi lk/mL] for milk samples. Linear regression analyses were performed to find the regression equation y=mx+b. The slope m was used to quantitate the total antioxidant capacity (TAC) , represented by the O R A C F L value. The O R A C F L value of a sample was calculated by dividing the slope of sample (ms) by the slope of Trolox Calibrator (m-rc)- The O R A C F L value for human milk was expressed as [umol Trolox/mL milk]. 29 2.2.3b Linearity and Precision of the O R A C F L Assay The linearity and precision of the O R A C F L assay was calculated using 10 independent runs of Trolox Calibrator based on the determination coefficient (R 2 ) and the coefficient of variation (%CV) , respectively. This data was obtained from the equations derived in linear regression analysis of the plotted net A U C and Trolox concentration of Trolox Calibrator. 2.2.3c Human Milk Dilution For the analysis of human milk, preliminary analyses were performed to find the optimal mi lk dilution. Whole human milk was diluted to 150x, 200x, 225x, 250x, 300x, and 350x. The O R A C F L assay was run for each dilution according to Table 2.2, which also provides the specific amounts of all chemicals needed to analyze milk samples. Table 2.2 Chemicals needed to analyze Trolox Calibrator and human milk samples 1 20uM Trolox (uL) 75nM PB 200nM FSS 60mM AAPH (uL) (UL) (UL) Blank 0 200 0 0 Control 0 140 60 0 Trolox calibrator 0 100 60 40 5 95 60 40 10 90 60 40 20 80 60 40 30 70 60 40 40 60 60 40 Sample Milk* (uL) 5 95 60 40 10 90 60 40 20 80 60 40 30 70 60 40 35 65 60 40 40 60 60 40 1 P B = Phosphate buffer FSS = Fluorescein sodium salt A A P H = 2 , 2 ' - azobis(2-amidinoproprane) di-hydrochloride * Diluted to: 150x, 200x, 225x, 250x, 300x, and 350x. 30 The steps described in Section 2.2.3a were followed, and the optimal dilution of human milk was chosen depending on a) Obtaining milk net A U C values within the working ranges of net A U C values for the Trolox Calibrator, b) Establishing maximal linearity in the plot of milk net A U C values vs milk concentration (referring to R 2=1.0 of the regression equation) and c) Finding the widest range of milk net A U C values that retain precision. 2.2.3d Antioxidant Recovery Experiments In order to complete the validation of the O R A C F L assay for human milk, unknown and known concentrations of Trolox standard were added to human milk to determine recovery and establish accuracy of the assay for measuring the oxidative stability of milk. In the first experiment, O R A C F L values were measured from one non-spiked milk sample and three milk samples spiked with 3 different unknown concentrations of Trolox. A l l four milk samples were from the same batch and were initially diluted to 400x. Volumes of 300uL of Trolox with concentrations of X , 2X , or 4 X u M were added to the same volume of diluted human milk, as shown in Table 2.3. Therefore, relative concentrations of Trolox added to milk were ViX, X , and 2 X u M , respectively; in which the concentration of X is the double of ' / 2 X and that of 2 X is the double of X . Table 2.3 Composition of spiked human milk samples 1 Milk sample Human milk X uM Trolox 2X uM Trolox 4X uM Trolox (UL) (UL) (UL) (UL) A 300 0 0 0 A + y2x 300 300 0 0 A+ X 300 0 300 0 A + 2 X 1 A •, , 300 0 0 300 A = non-spiked mother's mi lk VzX, X , 2 X represent relative amounts of Trolox standard added to milk P B = Phosphate buffer FSS = Fluorescein sodium salt A A P H = 2 , 2 ' - azobis(2-amidinoproprane) di-hydrochloride The four mi lk samples, non-spiked and spiked, were analyzed following the steps of the O R A C F L assay as summarized in Section 2.2.3a and Table 2.2. Linear slopes of the regression equations of mi lk samples, non-spiked or spiked, were used to calculate respective 31 Trolox concentrations (umol Trolox/mL milk). The recovery of Trolox standard added to milk was calculated by subtracting the total concentration of the non-spiked sample from each spiked milk sample (E.g.[A + y2 X] - [A]) . Sample (A + XA X ) was taken as a reference to determine the "theoretical" concentrations for samples (A + X ) and (A + 2X). The experimental Trolox standard concentration was compared to the "theoretical" concentrations and percent of accuracy was reported. The second experiment was designed to evaluate the accuracy of the O R A C F L method developed for mother's milk by adding a known concentration of Trolox standard to a human milk sample. A known volume of 20uM Trolox ranging from 5 to 30uL was added to lOuL of milk (Table 2.4). M i l k samples were spiked with Trolox standard and then analyzed for T A C following the the O R A C F L methodology as described in Section 2.2.3a and Table 2.2. Accuracy was determined by subtracting the net A U C value of the non-spiked milk sample obtained with lOuL ( A U C N S ) from the net A U C values of the spiked sample (AUCs). The resulted values corresponded to the experimental net A U C values of 2 0 u M Trolox added ( A U C T E ) - Net A U C T E values were compared to net A U C values of Trolox Calibrator ( A U C T C ) and percentage of accuracy was calculated. Table 2.4 Composition of milk spiked with known concentration of Trolox standard 1 Human milk Added 20uM 75nM PB 200nM FSS 60mM AAPH (UL) Trolox (uJL) (ML) (ML) (ML) 10 5 85 60 40 10 10 80 60 40 10 20 70 60 40 10 30 60 60 40 1 P B = Phosphate buffer FSS = Fluorescein sodium salt A A P H = 2 , 2 ' - azobis(2-amidinoproprane) di-hydrochloride 2.2.3e Antioxidant Capacity of Human Milk The validity of the O R A C F L assay to measure the antioxidant capacity of human milk was determined by comparing the standard curve of the Trolox Calibrator to the curve generated from the addition of human milk to the Trolox Calibrator. A n increase in the A U C values of the curve of Trolox Calibrator in which human milk was added would confirm the 32 presence of antioxidant activity in milk. A milk volume of 15uL was added to concentrations of Trolox Calibrator ranging from 0 to 40uL as shown in Table 2.5. Table 2.5 Composition of Trolox Calibrator spiked with a known volume of human milk. 2 0 u M Tro lox Added 7 5 n M P B 2 0 0 n M F S S 6 0 m M A A P H (ML) human mi lk (ML) (ML) (ML) (ML) 0 15 85 60 40 5 15 80 60 40 10 15 75 60 40 20 15 65 60 40 30 15 55 60 40 40 15 45 60 40 1 P B = Phosphate buffer FSS = Fluorescein sodium salt A A P H = 2 , 2 ' - azobis(2-amidinoproprane) di-hydrochloride 2.2.4 Data Analys is Descriptive statistics such as mean and standard deviation (±SD) were calculated for all raw data obtained. Regression analyses were performed to evaluate the linearity of Trolox Calibrator and the linearity of curves with different mi lk dilution factors. Coefficient of variation (%CV) was calculated to evaluate the precision of the O R A C F L assay. Regression analyses were also used to determine the regression equation required to calculate milk T A C values. Data transformation, regression analyses and calculations were performed using Microsoft® Excel 2002. Analyses were done in triplicate for all standard and milk sample concentrations, although otherwise specified. 33 2.3 R E S U L T S 2.3.1 The O R A C F L Assay Figure 2.1 shows a typical O R A C F L decay curve for Trolox Calibrator, in a Trolox concentration ranging from 0.0 to 4.0uM. With increased Trolox concentration, greater time is required for the fluorescence intensity to decay and approach zero. Figure 2.1 Effect of concentration of Trolox on Trolox Calibrator decay curve BO 10 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 Trolox uM/mL Figure 2.2 Standard curve for Trolox Calibrator after data transformation ( R 2 = 0.993) 34 Figure 2.2 shows the plot of Trolox Calibrator after data transformation. This represents the transformed data from Figure 2.1, showing the net area under the curve (AUC) versus the concentration of the Trolox standard. A l l Trolox Calibrator curves were done with six Trolox standard solutions with concentrations of 0.0, 0.5, 1.0, 2.0, 3.0, and 4.0uM. 2.3.2 L inear i ty and Precis ion of the O R A C F L Assay The linearity of the O R A C F L assay was determined by regression analysis by evaluating the slope and determination coefficient (R 2 ) of Trolox Calibrator. Regression equations were calculated according to y= mx + b, where y represents the net A U C obtained after data transformation (the integration of fluorescence readings from time 0 to 60min), m is the slope, x is Trolox concentration ((imol/mL), and b is the intercept. The results of regression analyses of 10 independent runs are summarized in Table 2.6. Linearity of the assay resulted in R = 0.990 with a variance of 2.2%. Table 2.6 Results from linear regression analysis of Trolox Calibrator R u n Slope Intercept R 2 1 8567.3 + 13.6 0.987 2 8596.7 + 13.5 0.992 3 8310.6 + 11.5 0.985 4 8509.1 + 11.2 0.984 5 • 8774.3 + 12.6 0.984 6 8669.7 + 11.8 0.996 7 8734.3 + 14.4 0.999 8 8791.0 + 11.6 0.991 9 8279.4 + 12.2 0.995 10 8746.6 + 11.74 0.993 Average a 8597.9 ± 184.7 (2.2) 12.4+1.1 (8.8) 0.990 ± 0.005 (0.5) Mean ± SD (%CV) 35 2.3.3 Human Milk Dilution To determine the optimal dilution of human milk samples for conducting T A C measurements, it was important to determine a range of net A U C values of a milk sample that were within those of Trolox Calibrator, over a concentration range of 0.0 to 4.0uM Trolox. The average net A U C values of 3 independent runs of the Trolox Calibrator are presented in Table 2.7, ranging from 10.6 to 45.9. On the other hand, the net A U C values for diluted human milk are presented in Table 2.8. Six different mi lk dilutions ranging from 150x, 200x, 225x, 250x, 300x to 350x were analyzed. The optimal dilution was determined in which: (a) M i l k net A U C fall within reproducible net A U C values for Trolox Calibrator, (b) A maximal regression coefficient demonstrating optimal linearity (R2=1.0) was obtained, and (c) A wide range of net A U C values from human milk were found. With these considerations, it was found that all dilutions fell within the net A U C values of the Trolox Calibrator, except for one (e.g. 350x). Two dilutions, 200x and 250x, gave the best result in linearity (e.g. R 2>0.990). Figure 2.3 represents the relationship between net A U C values of human milk and milk concentration of the two best dilutions. Dilution of mother's mi lk at 225x resulted in R 2=0.998 compared to the dilution of 200x which was R M X 9 9 5 . However, since dilution 200x resulted in a wider range of A U C values (from 12.6 to 31.8) than those of 225x (12.9 to 27.6), the former dilution was chosen for further human milk O R A C F L analyses. Table 2.7 Average net A U C values of Trolox Calibrator (n=3) uL 20uM Trolox Trolox (umol/mL) AUC value a 0 0.0000 10.6 ±0 .3 5 0.0005 17.6 ±0 .3 10 0.0010 21.3 ±0 .2 20 0.0020 32.8 ± 0 . 4 30 0.0030 39.5 ±0 .7 40 0.0040 45.9 ± 1.9 Mean ± SD 36 Table 2.8 Average net A U C values and R of different human milk dilutions A U C values Diluted milk j iL 150x 200x 225x 250x 300x 350x 5 15.2 ±0 .7 12.6 ±0 .7 12.9 ±0.3 11.8 ± 0 . 4 11.5 ± 0.1 10.2 ±0.5 10 17.8 ± 0 . 0 15.6 ±0 .9 14.4 ± 1.4 15.7 ± 0 . 0 14.2 ±0 .2 12.9 ± 1.3 15 22.2 ±0 .8 18.9 ±0 .3 17.2 ±1 .1 16.0 ± 1.7 15.6 ±0 .6 14.0 ±0 .1 20 25.2 ±0 .2 21.4 ±0 .9 19.5 ±0 .6 19.3 ±0 .8 18.1 ±0 .7 16.7 ± 1.2 30 28.1 ± 1.1 26.3 ± 1.5 23.8 ± 1.5 23.0 ±0 .8 21.7±0.5 18.9 ±0 .9 40 33.8 ± 2 . 4 31.8 ±0 .0 27.6 ±2 .0 26.7 ± 0 . 4 23.9 ±0 .4 22.3 ± 1.8 0.993 ± 0.006 0.995 ± 0.005 0.998 ± 0.007 0.983 ± 0.005 0.985 ± 0.005 0.985 ± 0.007 Mean ± SD Figure 2.3 Net A U C values of human milk and milk concentration at dilutions 200x (R 2=0.995) and 225x (R 2=0.998). 37 2.3.4 Ant iox idant Recovery Exper iments Exper iments o n recovery o f T r o l o x standard were performed to evaluate the increase on net A U C values and the accuracy o f the assay after m i l k samples were spiked w i t h T r o l o x standard. In the first experiment, m i l k samples were sp iked w i t h three different concentrations o f T r o l o x standard (e.g. / 4 X , X and 2 X T r o l o x ) . The in i t i a l concentration o f T r o l o x was unknown . F igure 2.4 represents the net A U C values o f sp iked m i l k and m i l k concentration. It can be observed that increasing the concentrat ion o f T r o l o x standard added to m i l k resulted i n l inear increases i n corresponding net A U C values. 40 35 30 A U C 25 20 15 10 X ^ m r * •—' .— • • A + 1 / 2 X A A + X x A + 2X 0 0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 Milk concentration (mL milk/mL) Figure 2.4 N e t A U C vs m i l k concentration o f human m i l k sp iked w i t h T r o l o x standard The O R A C F L value was calculated for a l l four m i l k samples. The experimental concentration o f T r o l o x standard added to the m i l k samples, i n u m o l T r o l o x / m L m i l k , was calculated by subtracting the T r o l o x concentration o f non-sp iked m i l k f rom each spiked m i l k sample. T h e exper imenta l concentrat ion o f T r o l o x added to the m i l k samples was 0.80, 1.57, and 2.91 u m o l T r o l o x / m L m i l k for I/2X, X and 2 X T r o l o x , respectively. Taken as reference value the concentrat ion o f V2X, the accuracy was calculated for X and 2 X w i t h values o f 98.1 and 90 .9%, respect ively (Table 2.9). 38 Table 2.9 Accuracy on experimental calculation of concentration of Trolox added to human milk 1 Concentration (umol Trolox/mL) A+ ViX A + X A + 2X Milk 3.50 ±0.09 4.31 ±0.19 5.08 ±0.11 6.41 ±0.01 [Spiked] - [non-spiked]a - 0.80 1.57 2.91 "Theoretical" Trolox added - Reference 1.60 3.20 Accuracy % Reference 98.1 90.9 X = Trolox standard added to human milk A = non-spiked human milk a Subtraction using the average concentration. In the second experiment on recovery of Trolox standard, a volume of lOuL of milk sample was spiked with a concentration of 20uM Trolox standard at increasing volumes from 5 to 30uL of Trolox. As shown in Figure 2.5, increased net A U C values were found in spiked milk samples when compared to its respective non-spiked milk sample. The horizontal line represents the net A U C values of lOuL of non-spiked human milk. . - - A A. ' ' ' ' , .4- " . A • ' A""'' • Milk A Spiked Milk 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 Trolox concentration (umol Trolox /mL) Figure 2.5 Net A U C values of non-spiked and Trolox spiked human milk. 39 Experimental net A U C values of Trolox added ( A U C T E ) were calculated from subtraction of the average net A U C values of non-spiked milk sample resulted at 10uL ( A U C N S = 13.7) from the net A U C values of the spiked sample (AUCs) . A U C T E was compared to the Trolox Calibrator net A U C values ( A U C T c ) , which used the same Trolox standard for spiking purposes (20uM Trolox). The accuracy in this experiment ranged from 91.7 to 97.9%, and it was different for different volumes of Trolox standard added to the milk sample. The average accuracy was 94.8 ± 3.2% (Table 2.10). Table 2.10 Accuracy on experimental calculation of concentration of Trolox added to a human milk sample 1 Trolox added (umol Trolox/mL) Average AUCs Average A U C T E A Average A U C T C B Accuracy % 0.0005 31.1 ± 0 . 1 17.3 ±0 .1 17.9 ±0 .2 97.1 ±0 .7 0.0010 35.1 ±0 .2 21.4 ±0 .2 21.8 ±0 .3 97.9 ±0.8 0.0020 43.4 ±0 .2 29.6 ± 0.2 32.1 ±0 .2 92.3 ±0.5 0.0030 49.2 ± 0.6 35.5 ±0 .6 38.7 ± 0 . 4 91.7 ± 1.5 Average 94.8 ±3 .2 'Average A U C N S = 13.7 A U C N S = area under the curve of non-spiked milk A U C s = area under the curve of spiked milk A U C T E = area under the curve of Trolox experimental A U C T E = area under the curve of Trolox Calibrator A A U C T E = A U C S - A U C N S B A U T T C from Trolox Calibrator 0.0005 to 0.0030 umol Trolox/mL. 2.3.5 Antioxidant Capacity of Human Milk A n increase in net A U C values when a known volume of human milk was added to the Trolox Calibrator confirmed the validity of the O R A C F L to measure the antioxidant capacity of human milk (Figure 2.6). 40 50 45 40 35 AUC 30 25 20 15 10 • Trolox Calibrator • Trolox + milk 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 Trolox concentration (|jmol/mL) Figure 2.6 Increase on net AUC values while adding human milk to Trolox Calibrator 41 2.4 DISCUSSION The O R A C assay using fluorescein, a photostable fluorescent probe ( O R A C F L ) , is an inexpensive, sensitive, and specific assay for evaluating or measuring antioxidant activity (Ou et al., 2001). O R A C F L values are based on the standard Calibrator curve from Trolox, a water-soluble vitamin E analogue. The O R A C F L assay has been widely used to assess antioxidant capacity o f food, dietary supplements, juices and wines, and biological samples, such as plasma and urine (Prior et al., 2003; Davalos et al., 2004; W u et al., 2004; Fernandez-Pachon, et al., 2005). Only one study has published the use of the O R A C assay using B -PE as the fluorescent probe to assess the antioxidant capacity of human milk (Alberti-Fidanza et al., 2002); however, no data was given on the standardization and validation of the assay. In this experiment, the O R A C F L assay was standardized and validated for human milk, and the resulting O R A C F L value, expressed in umol Trolox/mL milk, reflected total oxidative stability of human milk by measuring total antioxidant scavenging activity against a thermal-induced peroxyl radical attack. In order to calculate the O R A C F L value, fluorescein decay curves from milk samples and Trolox Calibrator were transformed into net area under the curve ( A U C ) values, expressed according to the respective concentration of milk or Trolox standard. After the fluorescein decay curve was transformed, a linear regression analysis was used to evaluate the linearity and precision of Trolox Calibrator and determine the optimal mi lk dilution factor required for the measurement of oxidative stability using the O R A C F L assay. A n important feature of linear regression analysis is the slope (m), which is required to calculate the O R A C F L value of a sample (dividing the slope of sample by the slope of Trolox Calibrator). The O R A C F L assay for human milk had satisfactory results since linearity had a •y coefficient R z > 0.990 and precision was within 15%, as established by Ou et al. (2001). The regression analysis for 10 independent runs of Trolox Calibrator, from concentrations 'y that ranged from 0.0 to 4 .0uM, demonstrated a linearity of R - 0.990 ± 0.005 and a precision of 2.2% between net A U C values and concentration of Trolox standard. Davalos et al. (2004) found similar precision results 1.9, 2.9 and 1.7% for three Trolox Calibrator curves. Determination of the optimal milk dilution factor was very important in order to be able to accurately calculate the O R A C F L value of mi lk samples with unknown antioxidant capacity. In this experiment, great efforts were taken to ensure that corresponding net A U C 42 values of mi lk samples would fall within the range of those from the standard Trolox Calibrator. Six different dilution factors were analyzed and by considering linearity of the data and the range of A U C values, a dilution factor of 200x was chosen for routine human milk O R A C F L analyses. Calculation of the O R A C F L value, which represents the total antioxidant capacity (TAC) of samples, requires the slope (m) from the regression equation of both the sample and the standard Trolox Calibrator. The O R A C F L value of human milk samples, expressed in umol Trolox/mL milk, was calculated by dividing the slope of the sample (ms) by the slope of Trolox Calibrator (mTc) ( O R A C F L value = ms I mjc). The O R A C F L assay was standardized and validated for measuring the antioxidant capacity of human milk, a reflection of milk oxidative stability. The antioxidant recovery of Trolox was demonstrated by two experiments. A n added Trolox standard or an increase in the concentration of Trolox standard added to milk samples produced increased net A U C values at the same mi lk concentration. In both recovery experiments, the accuracy of the method was calculated to be 95.4%. Recovery experiments were mainly used to verify the recovery of a Trolox standard after treatment and experimental analysis. The O R A C F L assay has been validated through linearity, precision, and accuracy in biological and food samples (Ou et al., 2001); however, the present study reports the validation of the O R A C F L assay in human milk samples. A n increase in the net A U C values reflects higher protection of chain-breaking antioxidants against AAPH- induced peroxyl radical damage (Cao et al., 1993; Ou et a l , 2001; Prior et al., 2003; Davalos et al., 2004). The last experiment showed the standard curve of Trolox Calibrator and the resulted curve after adding a known volume of human milk to the same Trolox Calibrator curve. Increased net A U C values were obtained reflecting a potential antioxidant capacity of human milk that may be due to natural milk chain-breaking antioxidants such as vitamin E. Different compounds in mi lk have been reported to possess antioxidant activity. Examples include the antioxidant vitamins A , E, C , and carotenoids, the peptides lactoferrin and casein, and the enzymes catalase, superoxide dismutase, and glutathione peroxidase (L 'Abbe and Friel, 2000; Lindmark-Mansson and Akesson, 2000; Hamosh, 2001; Friel et al., 2002; Shoji et a l , 2004; Kasapovic et al., 2005; Kitts, 2005; Shoji et al., 2005, Wong et al., 2006). The individual contribution of each and all of the potential 43 antioxidant components to the observed oxidative stability of human milk could not be determined in this study. 2.5 CONCLUSION The O R A C F L assay was standardized and validated for human milk to enable the measurement of total antioxidant capacity (TAC) against thermal-induced peroxyl radical attack, thus reflecting total oxidative stability of milk. The T A C of human milk was evaluated at different dilution factors and a factor of 200x was chosen for routine O R A C F L analyses. The O R A C F L assay was validated through linearity, precision, and accuracy. The standardized assay resulted in good linearity and precision. Accuracy values of above 90% were found with recovery experiments, after recovering Trolox standard added to human milk samples. This result reflected the accuracy of the O R A C F L assay after sample dilution, experimental analysis, data transformation, and computational calculations. In addition, the potential antioxidant protection of human milk was identified and may be due to natural chain-breaking compounds such as vitamin E or additional milk components showing antioxidant protection such as antioxidant vitamins, bioactive peptides, and the antioxidant enzyme matrix. 44 CHAPTER 3: A STUDY ON THE TOTAL ANTIOXIDANT CAPACITY, VITAMIN A AND E, AND FATTY ACID CONTENT OF H U M A N M I L K 3.1 INTRODUCTION Human milk is an important food for the growth and development of the breastfed infant which provides nutrients to meet nutritional requirements of the baby (Lawrence, 1994; Picciano, 1998; Malacarne et al., 2002; Newton, 2004). In addition, human milk provides extra-nutritional constituents which increase the importance of this physiological fluid for neonatal health. For example, the antioxidant defense system of human milk, which protects the breastfed infant from oxidative stressors, comprises of numerous bioactive components found to have antioxidant activity. Antioxidant vitamins (e.g. A , E, and C) ; bioactive peptides (e.g. casein and lactoferrin); and endogenous enzymes (e.g. catalase, superoxide dismutase, and glutathione peroxidase) (L 'Abbe and Friel , 2000; Lindmark-Mansson and Akesson, 2000; Hamosh, 2001; Friel et al., 2002; Shoji et al., 2004; Kasapovic et al., 2005; Kitts, 2005; Shoji et al., 2005, Wong et al., 2006) represent examples of the complex mixture of antioxidant components present in human milk. The total antioxidant capacity (TAC) of both food and biological systems has been studied often due to the potential interaction of different compounds that possess pro-oxidant and antioxidant activities. A better antioxidant protection against reactive oxygen species (ROS) was found in human milk compared to infant formula (Friel et al., 2002; Aycicek et al., 2006). Human milk T A C has been related to maternal intakes of dietary antioxidant vitamins. One study reported a positive relationship between the intake of antioxidant vitamins during pregnancy and lactation and the antioxidant capacity of human milk (Alberti-Fidanza et al., 2002). In addition, maternal antioxidant status has also been positively correlated with human milk T A C using the A B T S radical scavenging assay (VanderJagt et al., 2001). On the other hand, the pro-oxidant potential of individual human milk constituents has been investigated in terms of l ipid peroxidation reactions that can lead to oxidative stress (Granot et al., 1999; Miranda et al., 2004; Turoli et al., 2004). The effect of P U F A on the oxidative status of human milk was measured by assays that reflect products of l ipid 45 peroxidation (e.g T B A R S ) . A higher oxidation in human milk has been attributed to relative higher concentrations of L C - P U F A (Granot et al., 1999; Turol i et al., 2004) and to the activity of the enzyme bile salt stimulated lipase (BSSL) , which is not present in formula (Turoli et al., 2004). Interestingly, high levels of l ipid peroxidation products did not correspond to a lower T A C of human milk, suggesting that P U F A were not strictly related to milk T A C (Turoli et al., 2004). In this experiment, mature human milk was evaluated in terms of composition of antioxidant vitamins A and E, and individual fatty acids. The total fatty acid unsaturation index (UI) and milk T A C value were also determined. The purpose of this experiment was to elucidate potential relationships between total antioxidant capacity (TAC) of human milk and the content of antioxidant vitamins A and E and individual polyunsaturated fatty acids. The effect of total fatty acid unsaturation in milk was also evaluated in regard to milk T A C . 3.2 MATERIAL AND METHODS 3.2.1 Human Milk Samples A total of 77 human milk samples obtained for this study represent a convenience sample provided by the Chi ld & Family Research Institute (CFRI Vancouver, B.C.) as part of a larger study, PI Dr. Sheila Innis. 3.2.2 Collection of Milk Samples Human milk samples were collected during 2004-2005 from 77 different breastfeeding women (participants) at month 1 postpartum (mature milk). Participants were instructed to collect a milk sample expressed after 5 min of nursing (hindmilk). M i l k samples were frozen in the home freezer of participants following expression and collection. Samples were transported on ice to the CFRI laboratory within 24 hours after expression. Samples were stored at -80°C until analysis. 3.2.3 Background of Participants Potential study participants were identified from the registrations for low-risk delivery at the B C Women's Hospital and through advertising. El igible participants were 20-46 40 years of age, expected to deliver a single full-term (37-42 weeks) infant, and able to follow all study instructions. Exclusion criteria included any co-morbid condition such as diabetes, cardiac and renal diseases, tuberculosis or H I V / A I D S . Mothers who had known substance abuse and who had insufficient skills in English to read and understand the study were also excluded. In addition, some women took D H A supplements during gestation, but D H A intakes were not controlled and compliance could not be proven (See Section C. 1.2 of Appendix C) . 3.2.4 Demographics of Participants Demographic data of participants was provided by the C F R I team. Information such as maternal age, weight gain, ethnicity, educational level, income, and smoking habits were recorded. 3.2.5 Gestational Dietary Intakes of Participants Data on maternal dietary intakes was merely used as a descriptor o f the study participants. Maternal dietary intakes and the use of pre-natal multivitamin supplementation during gestation were obtained from a sub-sample of 29 women and was provided by the CFRI team. Food frequency questionnaires (FFQ) administered at 16 and 36 weeks of gestation were analyzed and dietary intakes from foods and pre-natal supplements were calculated as described in Appendix A . 3.2.6 Chemical Analyses of Human Milk Human milk samples were stored frozen at -80°C at the C F R I laboratory until analysis. Samples were thawed in cold water and either analyzed for fatty acids at the CFRI laboratory or transported to the Food Science Laboratory at the University of British Columbia for further analyses. Aliquots of human milk samples transported to the Food Science Laboratory were collected into sterile containers and kept on ice during their transportation. Upon arrival, samples were kept frozen at -80°C until further chemical analyses on vitamin A and E content and on total antioxidant capacity (TAC) . 47 3.2.6a Fatty Acid Analysis Analysis of individual human milk fatty acids was performed at the CFRJ laboratory. Data on fatty acid composition of milk samples (n=77) was provided by the C F R I team. Frozen milk samples were thawed in cold water and vortexed to prevent phase separation. A lOOul aliquot of human milk and 500ug of C17:0 internal standard (equivalent to about 600ug of lipid) were methylated with 2mL methanol-benzene (4/1, v/v) by modification of the direct transesterification method of LePage and Roy (1986). While vortexing, 200ul acetyl-chloride was added. Then, the tubes were capped tightly and heated at 100°C for 30min with vortex intervals at every 5min. The samples were neutralized by the addition of 6mL saline. Methyl esters were recovered by two extractions with 3mL pentane. The pooled pentane layers were dried under nitrogen gas and stored at -70°C until gas-liquid chromatography ( G L C ) analysis. Fatty acid methyl esters were separated on a 30m x 0.25mm ID, 0.20um f i lm non-blonded, fused sil ica capillary SP 2330 columns (Supelco, Belefonte PA) . Helium was used as the carrier gas, at a column flow of 1 mL/min and inlet pressure of 15 pounds per square inch. The inlet splitter was set at 10 to 1. Samples were injected at 80°C and kept constant for 2min; then increased to 170°C at a rate of 20°C/min. Temperature was held for 25min, then rose to 195°C at a rate of 20°C/min and held for 20min before subsequent analysis. The injectors and detectors were set at 240°C and 260°C, respectively. Fatty acid methyl esters were identified by comparison of retention times with those of authentic standards. 3.2.6b Vitamin A and E Analysis Vitamin A (all-^ra/«-retinol) and isomers of vitamin E (a-tocopherol, 8-tocopherol, and y-tocopherol) present in human milk were analyzed by liquid chromatography (HPLC) . M i l k samples required pre-treatment by fat extraction as described in Section B.3 of Appendix B, before H P L C analysis. H P L C parameters used are described in Section B.4 and in Table B . l of Appendix B. Vitamins A and E contents in human milk were calculated using regression equations of calibration curves of known standards and the vitamin E activity in human milk was calculated as described in Section B.5 of Appendix B. M i l k samples with resulting peak area [mAU*s] that did not fall within the range of the calibration curves of external standards could not be considered for the study. 48 3.2.6c Total Antioxidant Capacity Analysis Total antioxidant capacity (TAC) in human milk was analyzed by the O R A C F L assay as described in Chapter 2 (Section 2.2.3a), and by using specific chemicals and volumes shown in Table 2.2. A l l mi lk samples were diluted to 200x from original volume and analyzed in triplicate. M i l k T A C was represented by the O R A C F L value, and expressed as umol Trolox/mL milk. M i l k samples with net A U C values that did not fall within the range net A U C values of Trolox Calibrator could not be considered for the study. 3.3 DATA ANALYSIS 3.3.1 Screening Milk Samples A total of 17 human milk samples could not be used to calculate the milk T A C , due to milk net A U C values lower than those of Trolox Calibrator (Figure 3.1). It is noticeable that milk samples with undetectable T A C values also had low a-tocopherol contents. Therefore, data from a total of 60 human milk samples was selected for further statistical analyses, which represented 77.9% of the initial convenience sample. 5T i 1 0 1 2 3 4 5 6 Alpha-tocopherd ((jg r^rL m'lk) Figure 3.1 Relationship between milk T A C and a-tocopherol («=77). Data enclosed in circle denotes milk samples with undetectable T A C values («=17). 49 3.3.2 Evaluation of Milk TAC, Vitamins, and Fatty Acid Composition The relationships between milk T A C and vitamins A and E and individual fatty acid contents were determined by multiple regressions as described in Section C.2 of Appendix C. Independent variables considered were milk polyunsaturated fatty acids L A and A L A , and long chain polyunsaturated fatty acids ( L C - P U F A ) D H A , E P A , and A A , and contents of a-tocopherol and a l l - f r - t m s - r e t i n o l . Bivariate relationships were also analyzed. The relationships between the calculated milk unsaturation index (Ul) and measured milk T A C and vitamin A and E contents were also evaluated. U l was derived from the sum of the total number of unsaturation in mature human milk samples. This was calculated by multiplying the number of double bonds by the relative content of each unsaturated fatty acid present in milk as described in Formula (c.4) of Appendix C. 3.3.3 Evaluation of Human Milk Oxidative Stability The oxidative stability o f human milk was determined by mi lk T A C level. A high milk T A C corresponds to a high oxidative stability of human milk. The oxidative stability of human milk was evaluated by the effect of low, medium, and high levels of milk vitamins and fatty acids content. Distribution curves of milk variables were divided into tertiles showing low (I), medium (II) or high (III) levels as described in Appendix C (Section C.3.1). 3.3.4 Statistical Analysis Descriptive statistics such as mean, median, standard error of the mean (SEM), first and third quartiles ( Q l and Q3), and frequency distributions were calculated for all milk variables. The relationships among T A C values, vitamins A and E, and fatty acids contents in milk were evaluated using multiple linear regressions. In addition, Pearson correlations were used to evaluate bivariate relationships between milk T A C values and vitamins A and E, fatty acids, and U l . The effect of low, medium, and high levels of mi lk vitamins A and E and fatty acid contents on the oxidative stability of human milk were evaluated using odds ratios (OR). O R was calculated according to the Formula (c.3) given in Appendix C. A l l statistical analyses were done with SPSS© (version 10.0) and some calculations were performed in Microsoft® Excel 2002. The level of significance to detect statistical differences was set at p<0.05. 50 3.4 RESULTS 3.4.1 Demographic Characteristics Table 3.1 presents the demographic data obtained from the 60 participants who donated human milk samples at month 1 postnatally with detectable levels of milk T A C . The majority of the participants was Caucasian (73.3%) and had an annual household income higher than $50,000 (78.3%). A l l participants were well educated, having completed university or college. Two women reported smoking during gestation. The average maternal age at birth was 33.4 ± 0.5 years. Table 3.1 Demographic characteristics of participants (n=60) a Characteristic Maternal age at birth (years) 33.4 ±0 .5 ( 2 5 - 40) Pregnancy weight gain (kg) 14.9 ±0 .7 ( 3 - 3 1 ) Number of live births 1.6 ± 0.1 ( 1 - 4 ) (including present baby) # of participants (%) Ethnicity: Caucasian Chinese Other 44 7 9 (73.3) (11.7) (15.0) Educational level: College 16 44 (26.7) (73.3) University Household Income: < $20,000 $20,000 - 50,000 > $50,000 12 47 (1.7) (20.0) (78.3) Smoking N o Yes 58 (96.7) 2 (3.3) a Mean ± S E M , range in parenthesis 51 3.4.2 Maternal Dietary Intakes during Gestation Data on maternal dietary intakes during gestation was analyzed from a sub-sample of 29 participants and used only as a descriptor of the participants' diets. Statistical analysis comparing intakes at 16 and 36 weeks of gestation, showed no statistical difference, suggesting stable maternal dietary habits. Therefore, individual intakes of all participants were taken as average daily intakes from the two data values. Maternal macronutrient intakes are presented as percent of energy intake in Table 3.2. Average daily energy intake was 2206.02 ± 135.42 kcal/d. Dai ly protein intake represented 16.76 ± 0.41% of energy, carbohydrate 54.16 ± 1.01% of energy, and fat 31.68 ± 0.86% of energy. Looking at the average daily intakes of macronutrients in percent of energy, all were within the Acceptable Macronutrient Distribution Range ( A M D R ) . Table 3.2 Maternal average daily intakes of macronutrients given as percent of energy intake during gestation («= 29) a Macronutrient % of energy AMDR % b Protein 16.76 ±0.41 (12.11-23.61) 1 0 - 3 5 Carbohydrates 54.16 ± 1.01 (38.23-62.44) 4 5 - 6 5 Fat 31.68 ±0.86 (24.04-45.05) 2 0 - 3 5 a Mean ± S E M ; range in parenthesis b A M D R : Acceptable Macronutrient Distribution Range Dai ly intakes of some individual fatty acids are presented in Table 3.3. Average daily intakes of total fat was 78.18 ± 5.85 g/d, saturated fat (SFA) 27.69 ± 2.17 g/d, monounsaturated fat ( M U F A ) 28.43 ± 2.39 g/d, and polyunsaturated (PUFA) 11.20 ± 0.84 g/d. The average daily intake of linoleic acid (LA) was 9.12 ± 0.74 g/d (3.75 ± 0.23% total energy) and of linolenic acid ( A L A ) was 1.14 ± 0.08 g/d (0.48 ± 0.03% total energy). The average daily intakes of arachidonic acid (AA) , eicosapentaenoic acid (EPA) , and docosahexaenoic acid (DHA) were 0.12 ± 0.01, 0.09 ± 0.02, and 0.15 ± 0.03 g/d, respectively. Some women took D H A supplements during gestation; however, intakes from D H A supplements were not controlled for compliance. In addition, there was a large overlap 52 in D H A content in mi lk samples, at month 1 postnatally, from both women who mentioned taking and not taking D H A supplements (Appendix C.2). Table 3.3 Maternal daily intakes for individual fatty acids during gestation (n= 29) Fatty acids Average In take 8 M e d i a n Intake b L A (18:2 co-6) (g/d) 9.12 ±0.74 (3.39--19.53) 8.09 (6.41, 10.56) A L A (18:3 co-3) (g/d) 1.14 ±0.08 (0.49 -2 .41) 0.99 (0.83, 1.36) A A (20:4 co-6) (g/d) 0.12 ±0.01 (0.02 -0 .30) 0.10 (0.06,0.15) E P A (20:5 co-3) (g/d) 0.09 ± 0.02 (0.00 -0 .56) 0.04 (0.01,0.12) D H A (22:6 co-3) (g/d) 0.15 ±0.03 (0.02 - 0.64) 0.08 (0.04,0.24) a Mean ± S E M ; range in parenthesis. b M e d i a n intakes; quartile 1 and quartile 3 ( Q l , Q3) in parenthesis. Maternal average intakes of antioxidant vitamins are presented in Table 3.4. From the 29 participants, only one woman chose not to take a pre-natal supplement, but was recorded as taking iron supplements. Average daily intakes of antioxidant vitamins from food sources were 1303.28 ± 96.03ug/d, 8.68 ± 0.69mg/d, and 244.49 ± 14.61mg/d for vitamins A , E and C , respectively. The use of a pre-natal multivitamin supplement increased average daily intakes to 199.3%, 320.0%, and 152.9% for vitamins A , E and C, respectively. Table 3.4 Maternal average daily intake of antioxidant vitamins during gestation («= 29). Diet only Diet + V i t S " R D A / U L C Vitamin A 1303.28 ±96.03 2597.84 ±207.27 770 / 3000 ( R A E ug/d) (400.26 - 2209.89) (1202.93 -4781.88) Vi tamin E 8.68 ±0.69 21.78 ± 1.18 15/1000 (alpha-TE mg/d) (3.45-20.31) (11.45-38.91) Vi tamin C 244.49 ± 14.61 373.87 ±33.84 85 /2000 (mg/d) . (99.14-461.40) (199.14- 1190.53) Mean ± S E M ; range in parenthesis; V i tS : Vitamin Supplement; 0 R D A : Recommended Dietary Allowance / U L : Upper Level 53 3.4.3 Antioxidant Capacity, Vitamin, and Fatty Acid Characteristics of Human Milk The T A C value, vitamins A and E contents, and vitamin E activity of mature human milk are listed in Table 3.5. M i l k T A C represents the oxidative stability of human milk, in which a higher T A C value reflects higher oxidative stability, and thus potentially greater antioxidant protection for the breastfed infant. The average milk T A C was 3.41 ± 0.07umol Trolox/mL. O f the different tocopherol isomers measured in milk samples, the highest concentration was found to be for a-tocopherol (2.32 ± 0.1 lug /mL) , furthermore, it was the main component of the vitamin E activity (2.37 ± 0.12ug/mL). Vi tamin A content, measured as all-frara'-retinol, was 0.08 ± O.Olug/mL mature milk, with a relative large range among subjects (e.g. 20-fold difference). Table 3.5 T A C and vitamin content in mature human milk (n=60). Milk variable Average a Median b T A C (umol Trolox/mL) 3.41±0.07 (2.26 --4.90) 3.41 (3.03 --3.80) a-tocopherol (ug/mL) 2.32 ±0.11 (0.66 -- 5.02) 2.23 (1.79--2.93) 8-tocopherol (ug/mL) 0.11 ±0.01 (0.00 --0.56) 0.09 (0.06--0.12) y-tocopherol (ug/mL) 0.46 ± 0.03 (0.11 --1.27) 0.45 (0.29--0.59) Vitamin E activity (ug/mL) 2.37 ±0.12 (0.71 -- 5.07) 2.30 (1.82 -- 2.99) alWrcmy-retinol (ug/mL) 0.08 ±0.01 (0.01 -- 0.20) 0.09 (0.05 --0.11) a Mean ± S E M , range in parenthesis b Median, QI - Q3 range in parenthesis. This range represents 50% of human milk samples. Data on the individual fatty acids present in human milk, along with the calculated unsaturation index (UI) is given in Table 3.6. Contents of groups of fatty acids in g/lOOg F A were: total S F A 39.04 ± 0.67, total M U F A 38.89 ± 0.56, total P U F A 16.82 ± 0.48, omega-6 P U F A 14.74 ± 0.42, and omega-3 P U F A 2.06 ± 0.09. The content of D H A in mature milk was 0.30 ± 0.02g/100g F A ; however, a wide range was found 0.08- l . l l g / lOOg F A . The unsaturation index (UI) value was 77.5 ± 1.2g/100g F A , and it was calculated as the sum of all number of double bonds of unsaturated fatty acids multiplied by their relative content in milk. Fatty acids present in higher concentration were the saturated palmitic acid (C16:0), 54 the monounsaturated oleic acid (Cl8:1) , and the polyunsaturated linoleic acid (C18:2, co-6), with average concentrations of 20.31 ± 0.38, 35.29 ± 0.56, and 13.37 ± 0.41g/100g F A , respectively. The major contributors to UI were oleic acid and L A . Table 3.6 Fatty acid content and UI in mature human milk (g/lOOg F A ) («=60) Milk Fatty Acids Average a Saturated (SFA) Capric (C10:0) 0.87 ±0.04 ( 0 . 2 4 - 1.98) Laurie (C12:0) 4.93 ±0.24 ( 1 . 2 9 - 13.51) Myrist ic (C14:0) 6.17 ±0.25 ( 2 . 6 8 - 13.78) Palmitic (C16:0) 20.31 ±0.38 (14.71 - 26.20) Stearic (C 18:0) 6.43 ± 0.20 (4.03 - 14.76) Total S F A 39.04 ± 0.67 (27.47 --50.33) Monounsaturated ( M U F A ) Palmitoleic (C16: l ) 2.68 ±0.09 ( 1 . 4 0 - 4.43) Oleic (C18: l ) 35.29 ±0 .56 (21.00--45.55) Total M U F A 38.89 ±0 .56 (24.89--48.05) Polyunsaturated (PUFA) L A (C 18:2, co-6) 13.37 ±0.41 ( 8 . 4 9 - 22.82) A A (C20:4, co-6) 0.42 ±0 .01 ( 0 . 2 6 - 0.58) Total omega-6 (co-6) 14.74 ±0.42 ( 9 . 2 9 - 24.44) A L A (C l8 :3 , co-3) 1.52 ±0 .07 (0 .40- 3.41) E P A (C22:5, co-3) 0.09 ±0 .01 (0 .00- 0.38) D H A (C22:6, co-3) 0.30 ±0.02 (0 .08- 1.11) Total omega-3 (co-3) 2.06 ±0 .09 ( 0 . 9 0 - 3.98) Total P U F A 16.82 ±0.48 (10.67--28.44) U I b 77.5 ± 1.2 ( 5 8 . 0 - 9 9 . 7 ) a Mean ± S E M , range in parenthesis b Unsaturation Index: UI = Z (mj x rj) 55 Multiple regression analyses were used to evaluate the relationship between the independent variables: a-tocopherol, all-fr-ara-retinol, D H A , E P A , A A , A L A , A A , co-3 and co-6 F A content, and the dependent variable milk T A C . Results only suggested a relationship between milk T A C ( O R A C F L value) and a-tocopherol content (Appendix C , Section C.2.1). Coefficients obtained from bivariate linear regression analysis are given in Table C.7 of Appendix C , and demonstrated a positive relationship between mature human milk T A C and a-tocopherol content, at month 1 postnatally, r=0.439 (p<0.001) (Figure 3.2). The scatter plot showing milk T A C and alRra/is-retinol content (Figure 3.3) did not show any significant relationship. Other statistically significant correlations in human milk samples, obtained at month 1 postnatally, were found between a-tocopherol and all-fr-cms-retinol contents r=0.557 (p<0.001), mi lk D H A and E P A contents r=0.621 (p<0.001), D H A and A L A contents r=0.316 (p<0.05), and A L A and L A contents r=0.632 (pO.OOl) . Plots showing these significant relationships are presented in Section C.2.3 of Appendix C. 5.0 T , S ~ 4.5 2.0] 0 1 2 3 4 5 6 /^pha-tocopherd (pg/rri. rrilk) Figure 3.2 Relationship between milk T A C and a-tocopherol r=0.439 (pO.OOl) («=60). 56 5.0 4.5 x o o 4.0 o E 3.5 O 3.0 2.5 2.0 0.00 • M .05 .10 .15 all-trans-retinol ((jg/rrL) .20 Figure 3.3 Scatter plot to show milk T A C and all-^rara-retinol, r=0.155 (p=0.24) (n=60) No relationships were obtained between milk D H A content and T A C value (Figure 3.4), a-tocopherol, and alRra/w-retinol contents (Table C.7 Appendix C). 5.0 2* 4.5 x o 4.0 3.5 o E 3 3.0 o 2.5 2.0 J 0.0 % • • • • .2 1.0 1.2 .4 .6 .8 MlkDHMg/100gFA) Figure 3.4 Scatter plot to show milk D H A content and milk T A C , r=-0.123 (p=0.35) (n=60) 57 No relationships were obtained between the degree of mi lk fatty acid unsaturation, represented as UI (%) and milk T A C , at month 1 postnatally (Figure 3.5). 5.0i ^ 4.51 x _o o h= ~o < • « • 4 .0' 3.5. 3.0. 2.5. 2.0, 50 60 70 80 90 100 110 Mlk U (%) Figure 3.5 Scatter plot to show milk UI and milk T A C , r=-0.054 (p=0.68) (n=60) 3.4.4 Oxidat ive Stabi l i ty of H u m a n M i l k The oxidative stability of mature human milk, represented as milk T A C , was further evaluated using odds ratio (OR) analysis. Tertiles o f low (I), medium (II), and high (III) contents of vitamin A and E, individual fatty acids, and fatty acid unsaturation (UI) in milk were variables o f consideration. Ranges of tertiles of variables are given in Table 3.7. Table 3.8 presents the O R for having low milk T A C values and Table 3.9 presents the O R for having high milk T A C values. The interpretation of O R and 95% CI is described in Section C.3.3 of Appendix C. M i l k a-tocopherol was the only component that resulted in a statistically significant effect on milk oxidative stability. 58 Table 3.7 Tertiles of low (I), medium (II), and high (III) contents of milk variables a Tertiles Milk variable I II 111 M i l k T A C value (umol Trolox/mL) 2 .26 -3 .17 (n=22) 3.18 («= - 3 . 6 2 =16) 3 .66-4 .90 («=22) M i l k a-tocopherol (Ug /mL) 0 .66-1 .93 («=23) 1.97 («= - 2 . 6 4 =18) 2 .71 -5 .02 («=19) M i l k all-fr-am'-retinol (ug/mL) 0 .01 -0 .06 («=23) 0.07 - 0 . 0 9 =14) 0 .10-0 .20 («=23) M i l k P U F A (g/lOOg F A ) 10.67-15.06 («=24) 15.40 («= - 18.24 =17) 18.41-28.44 (n=19) M i l k L A (g/lOOg F A ) 8 .49-11.96 («=25) 12.10 (n-- 14.63 =19) 15.07-22.82 («=16) M i l k A L A (g/lOOg F A ) 0 .40 -1 .27 («=24) 1.29 («= - 1.75 =21) 1.82-3.41 ("=15) M i l k D H A (g/lOOg F A ) 0 .08-0.21 («=22) 0.22 («= - 0 . 3 6 =25) 0 .40-1.11 (n=\3) M i l k co-3 L C - P U F A (g/lOOg F A ) 0 .17-0.41 (w=22) 0.44 (n= - 0 . 6 4 =24) 0 .65-1 .74 («=14) M i l k co-6 L C - P U F A (g/lOOg F A ) 0 .73 -1 .17 («=21) 1.20 («= - 1 . 3 5 =17) 1.38-1.69 (n=22) M i l k UI (%) 58.03-73.25 (n=22) 74.47 (rr-- 8 1 . 1 6 =18) 81.54-99.69 («=20) Tertiles are given as range, and the number of participants on each tertile is in parenthesis. 59 Table 3.8 presents that a tertile having high content of a-tocopherol in milk resulted in a protective effect against low milk T A C . The range of a-tocopherol protection was from 18 to 94%, expressed by the confidence interval (95% CI = 0.06-0.82) (p=0.02). On the other hand, the O R for having high milk T A C values are presented in Table 3.9. A high content of a-tocopherol in milk (2.71-5.02(ig/mL) resulted in significantly increasing the odds for having high mi lk T A C values (p=0.02). A high content of all-frvms-retinol in mi lk (tertile III) could also be a potential component for increasing milk T A C values (95% CI= 0.75-6.29); but, as previously shown in Figure 3.3, the two milk variables did not shown any relationship at p<0.05. A s presented in Table 3.8, the effect of L A , A L A , D H A , co-3 L C - P U F A and co-6 L C -P U F A on decreasing milk T A C values was greater while the respective contents were higher, as shown with the comparison of tertiles I and III. However, while comparing medium and high contents (tertile II vs III) of D H A , co-6 L C - P U F A , and unsaturation index (Ul), the O R for low milk T A C values was indeed lower at a high content (tertile III). Medium contents of milk P U F A and A L A (tertile II), 95% CI 0.97-9.75 and 0.94-8.32, respectively, almost provided a positive effect on increasing milk T A C values. High contents of D H A in milk (0.40 - l . l l g / lOOg F A ) compared to low and medium contents (0.08-0.2 lg/ lOOg F A and 0.22-0.36g/100g F A , respectively), resulted in a smaller O R for having high milk T A C values (OR^O.72). Nevertheless, the true O R values for the three milk D H A tertiles, represented by the 95% CI, suggested similar effects both decreasing and increasing the odds for having high milk T A C . Similar results were found for U l , also suggesting lack of statistical effect in increasing milk oxidative stability. 60 Table 3.8 Odds ratio and 95% CI for having low milk T A C value M i l k var iable Terti les I II III 1.60 2.23 0.22 * M i l k a-tocopherol (0 .56-4.62) (0.74 - 6.80) (0.06-0.82) 1.60 1.46 0.46 M i l k all-fr-am'-retinol (0 .56-4.62) (0.43 - 4.62) (0.15-1.41) 0.79 0.92 1.40 M i l k P U F A (0.27-2.28) (0 .30-2.90) (0.41-4.21) 0.95 1.01 1.05 M i l k L A (0.33-2.73) (0 .34-3.06) (0.33-3.34) 0.79 0.80 1.75 M i l k A L A (0.27-2.28) (0 .27-2 .39) (0.55-5.60) 0.72 1.28 1.10 M i l k D H A (0.24-2.13) (0 .45-3.64) (0.33-3.77) 0.72 1.06 1.41 M i l k co-3 L C - P U F A (0.24-2.13) (0 .37-3.05) (0.43-4.62) 0.58 2.60 0.72 M i l k co-6 L C - P U F A (0.19-1.76) (0 .84-8 .06) (0.24-2.13) 0.98 1.60 0.90 M i l k U l (0 .34-2.87) (0 .53-4.86) (0.30-2.70) a 95% Confidence intervals are in parenthesis. * Statistically significant at p<0.05 61 Table 3.9 Odds ratio and 95% CI for having high milk T A C value M i l k var iable Terti les I II III 0.22 1.15 3 .75* M i l k a-tocopherol (0 .07-0.75) (0 .38-3 .50) (1 .22 - 11.55) 0.64 0.62 2.17 M i l k all-fr-ara-retinol ( 0 . 2 2 - 1.90) (0 .18-2.20) (0.75 - 6.29) 1.06 3.07 0.34 M i l k P U F A (0.37-3.05) (0.97 - 9.75) (0.10-1.16) 0.95 1.94 0.48 M i l k L A (0 .33-2.73) (0 .65-5.81) (0.14-1.67) 0.79 2.80 0.34 M i l k A L A (0.27-2.28) (0.94 - 8.32) (0.09-1.31) 1.33 0.95 0.72 M i l k D H A (0.46-3.87) (0 .33-2 .73) (0.20-2.56) 1.31 1.06 0.62 M i l k co-3 L C - P U F A (0 .46-3.87) (0 .37-3.05) (0.18-2.20) 1.50 0.64 0.98 M i l k co-6 L C - P U F A (0.51-4.40) (0 .20-2 .07) (0.34-2.87) 0.98 0.81 0.90 M i l k U l (0 .34-2.87) (0 .26-2.54) (0.30-2.70) a 95% Confidence intervals are in parenthesis. * Statistically significant at p<0.05 62 3.5 DISCUSSION 3.5.1 Participants: Background, Demographics, and Dietary Intakes during Gestation M i l k donors were women between 25-40 years of age. Only 2 women reported smoking during gestation; however, a lack of detailed information about their smoking habits could be an important l imiting factor in this study. Maternal dietary intakes during gestation were analyzed from a sub-sample of 29 women. Interpretation of these data indicated healthy dietary habits for all women; an expected finding, since women were educated and had high household incomes. The use of the Food Frequency Questionnaire (FFQ) to evaluate maternal dietary intakes during gestation, has the advantage to estimate the usual frequency of consumption of foods for a specific period of time. On the other hand, a major limitation is the inaccuracy on quantification of intakes; for example, incomplete listing of all possible foods can lead to errors in frequency estimation (Thompson and Byers, 1994). No significant differences were seen between maternal dietary intakes at 16 and at 36 weeks of gestation for both macro and micronutrients. This result suggested that gestational dietary intakes of these women were relatively stable; a finding similar to that previously reported in pregnant women (Elias and Innis, 2001). Maternal daily fat intakes during gestation were found to be similar to fat intakes of pregnant Canadian women reported by Innis and Elias (2003). Six participants had a total fat intake above the A M D R (36.18-45.05%) (data not shown). Dai ly intakes of fatty acids recommended by the International Society for the Study of Fatty Acids and Lipids ( ISSFAL) are 4.44g L A and 2.22g A L A , and 0.22g E P A for adults and >0.3g D H A for pregnant women (Simopoulos et al., 1999). In the present study, only 1 participant did not meet the recommendation for L A (4.44g/d) and 28 participants had intakes below 2.22g/d for A L A and only 5 participants (17.2%) met the I S S F A L recommendation for D H A > 0.3g/d (data not shown). These results indicated that women who participated in this study were not meeting the I S S F A L recommendations for intakes of D H A , as reported by Innis and Elias (2003). Pre-natal multivitamin supplementation was common for women during gestation, with only one participant not taking supplements. One pre-natal tablet provided a vitamin A dosage that is higher than the U L . Vitamin E from pre-natal supplements was an important practice since maternal diets did not provide an adequate supply; however, it would be 63 difficult to exhibit maternal or fetal deficiency since vitamin E concentration in maternal circulation is known to increase during gestation due to high l ipid mobilization (Oostenbrug et al., 1998; Ladipo, 2000). A l l participants had vitamin C intakes above the R D A (e.g. 85mg/d) even before taking the pre-natal supplement. There were no cases where intake was above the vitamin C upper level of 2000mg/d. Vitamin C supplementation is recommended during gestation because concentrations in maternal circulation are known to decrease around 50% due to fetal uptake (Ladipo, 2000). Overall, despite the healthy dietary habits of participants during gestation in this study, the absence of similar information on maternal dietary intakes during lactation and, moreover, at the actual time of sampling of human milk, made it impossible to relate participant dietary information to human milk antioxidant activity (e.g. mi lk T A C ) . 3.5.2 Ant iox idant Capaci ty , V i t am in , and Fatty A c i d Character ist ics of H u m a n M i l k There is limited information concerning human milk total antioxidant capacity (TAC) using the O R A C F L assay. Former studies have suggested a positive, time-dependant relationship between maternal antioxidant intake and milk T A C values (Alberti-Fidanza et al., 2002). In the present study, the T A C of human milk from Canadian women was measured by the O R A C F L assay, which in turn was used as a biomarker to evaluate milk oxidative stability. Due to lack of food frequency data at the time of sampling, at month 1 postnatally, it was not possible to relate maternal dietary intakes with mi lk T A C . Several studies have reported the effect of maternal dietary vitamin A and E intakes, stages of lactation, and vitamin supplementation on human milk vitamin contents (Jansson et al., 1981; A l i et al., 1986; Ortega et al., 1997; Ortega et al., 1999; Macias and Schweigert, 2001; Olafsdottir et al., 2001; Gossage et al, 2002; Meneses and Trugo, 2005; Korchazhkina et al., 2006). The significant positive relationship found between a-tocopherol and all-/nmy-retinol content in human milk, r=0.587 (pO.OOl), suggests a habitual dietary multivitamin supplementation pattern during pregnancy and lactation that could have contributed to oxidative stability of mother's milk; however, these findings are not conclusive. In addition, the positive relationships between milk D H A and E P A contents and milk L A and A L A contents, suggested maternal dietary intakes of fish and vegetables, respectively. 64 The T A C values of mature human milk were found to be variable among participants, ranging from 2.26—4.90umol Trolox/mL milk («=60). The T A C values of human milk (month 1 postpartum) studied herein were higher than those values reported by Alberti-Fidanza et al. (2002) obtained from the measurement of human milk at day 20 t h postpartum from Italian women, (range 0.36-2.18umol Trolox/mL milk). The difference in findings between these two studies could be due to the employment of different fluorescence probes used in the O R A C assay, which are required to measure the oxidative inhibition in the presence of a milk sample. The O R A C P E assay uses B-phycoercythrin (B-PE) as the fluorescent probe, which lacks of photostability in comparison to fluorescein (FL) (Ou et al., 2001), the fluorescent probe of the O R A C F L assay used in the present study. In addition, B-P E forms nonspecific protein binding complexes which result in variable reactivity (Cao and Prior, 1999) which can lead to falsely lower T A C values than those of the O R A C F L assay (Oueta l . , 2001). The average content of a-tocopherol in mature human milk obtained herein was 2.32 ± O. l l ug /mL milk, a similar finding to that reported in mature human milk from Spanish (Ortega et al.,1999) and Cuban women (Macias and Schweigert, 2001). On the other hand, y -tocopherol was the second major tocopherol isomer identified in human milk, as reported in other studies (Olafsdottir et al., 2001; Korchazhkina et al., 2006), which also constitutes a major vitamin E component of soybean and corn oils, walnuts, and pecans (Institute of Medicine, 2000; Jian et. al., 2001), as well as being the major vitamin E form obtained in the U S diet (e.g. 70% of total vitamin E consumed) (Jian et. al., 2001). In the present study, the average contents of y and 5 tocopherols in human milk compared to that of a-tocopherol were minimal, e.g. 19.8% and 4.7%, respectively. The vitamin E activity was also calculated and used to represent the biological activity of the isomers measured in human milk; the average value of vitamin E activity in the present study was 2.37 ± 0.12ug/mL milk. The content of a-tocopherol accounted for 97.9% of vitamin E activity in mature human milk, which agrees with previous reports (e.g. 94%) (Olafsdottir et al., 2001). In comparison to a-tocopherol, y-tocopherol has been reported to possess a stronger antioxidant activity and a higher efficiency against chronic illness such as prostate cancer (Helzlsouer et al., 2000), in addition to exhibiting a stronger in vitro antioxidant activity (Bramley et al., 2000). It may be possible that y-tocopherol provides the greatest potential antioxidant protection for the 65 breastfed infant; however, factors influencing its bioavailability or retention in the body need to be considered. For example, urinary excretion of y-tocopherol metabolites, but not of a-tocopherol, indicate that levels of a-tocopherol are maintained to a greater extent in the human body (Leonard et al., 2005). This finding is strengthen by the understanding that the specificity to the tocopherol binding protein (a-TTP) present in the liver is important for vitamin E isomers to exert biological effects (Schneider, 2005). The major vitamin A component in human milk is retinol, which is present as retinyl esters, accounts for 90% of vitamin A activity (Macias and Schweigert, 2001; Picciano, 2001); a finding that is related to the fact that retinol uptake by the mammary gland is more effective than the uptake of carotenoids (Meneses and Trugo, 2005). The vitamin A content of Canadian human milk measured herein was similar to aW-trans-retinol values previously reported from Icelandic milk (Olafsdottir, et al., 2001). Other studies have reported the content of main carotenoids in human milk (e.g. lycopene, P-carotene, and lutein) (Macias and Scheweigert, 2001; Gossage et al., 2002; Meneses and Trugo, 2005). The significance of this information to mi lk antioxidant status was not measured in the present study since carotenoid concentrations decreased rapidly over stage of lactation and storage (Macias and Schweigert, 2001). The O R A C F L procedure is based on the generation of peroxyl radicals by thermal decomposition of A A P H and the free radical scavenging activity of antioxidants present in human milk (Prior and Cao, 1999). According to the results obtained in this study, 19.3% of human milk oxidative stability was attributed to mi lk a-tocopherol content (r=0.439, p<0.001). Previous reports have found that the antioxidant activity of a-tocopherol is manifested against peroxyl radicals in food and biological systems, such as bovine milk and lipoproteins (Kitts, 1997; Ternay and Sorokin, 1997; Oostenbrug et al., 1998; Lindmark-Mansson and Akesson, 2000; Schneider, 2005). Similar to our results, X u et al. (2001) found that vitamin E in rice bran exhibited significant antioxidant activity against A A P H -induced cholesterol oxidation. The antioxidant activity in tomatoes, measured by the T E A C assay, was well correlated with ascorbic acid and lycopene, the major antioxidant components in tomato (Cano et al., 2003). A study evaluating the antioxidant capacity of infant formulas by the T E A C assay, reported a higher T A C in formulas containing higher vitamin A , E, and C contents (Turoli et al., 2004). Despite this evidence, showing a 66 significant relationship between these vitamins and T A C values, a statistically significant positive pattern was not found between vitamin A and T A C values of human milk in the present study. Human milk lipids were measured from hindmilk, the milk expressed after 5 min of nursing, and contains up to three times the l ipid content of foremilk, the milk expressed at the beginning of nursing (Lawrence, 1994; Picciano, 2001; Turoli et al., 2004). The composition of individual mi lk fatty acids (FA) was found to be similar to that reported in other studies (Boersma et al., 1991; Francois et al., 1998; Picciano, 2001; Malacarne et al., 2002). Lipids account for 50% of total energy for the breastfed infant. The major fatty acids in mature human milk studied herein were oleic acid, palmitic acid and linoleic acid, as determined in previous studies (Malacarne et al., 2002). The degree of fatty acid unsaturation was calculated with a formula that recognizes the unsaturation index (UI) of this biological fluid. Higher mi lk UI reflects a greater susceptibility for l ipid peroxidation, and eventual rancidity (Turoli et al., 2004), both of which are reflected by the oxidative stability of human milk. Results in this study showed no statistically significant effect of individual fatty acids on milk T A C values, nor milk vitamin A or E contents. Our results did not support findings from previous studies. For example, a potential antioxidant activity of omega-3 L C - P U F A in biological systems, such as human plasma (Wander, 2001; Barbosa et al., 2003; Mor i , 2004), was not found for human milk in this study. In addition, a trend showing a negative relationship between UI and a-tocopherol was reported in formula (Miquel et al., 2004), suggesting that the antioxidant role of this vitamin is important for contributing to the oxidative stability of unsaturated lipids in human milk. The degree of fatty acid unsaturation was not affecting milk antioxidants nor milk T A C , according to the results found herein. In the present study, a lack of effect between multiple independent milk variables, such as vitamin A and unsaturated fatty acids on human milk T A C , suggested that milk total antioxidant capacity (TAC) could represent a complex mixture of compounds present in milk, as previously described, that may all have some bearing on maintaining milk oxidative stability. 67 3.5.3 Oxidat ive Stabi l i ty of H u m a n M i l k Odds ratio analysis (OR) was used herein to determine potential milk components that could affect mi lk oxidative stability. Low, medium, and high contents of a, y and 8 tocopherols, all-frarcs-retinol, unsaturated fatty acids, and U l were the milk components under analysis. Medium contents of total P U F A and A L A in human milk (OR= 3.07 and OR= 2.80, respectively) resulted in potentially enhancing milk oxidative stability (Table 3.9), while high contents resulted in an opposite effect (OR= 0.34). These findings suggested that the higher the concentration of P U F A in milk, the greater the tendency for reactive oxygen species (ROS) interaction with milk components, especially L C - P U F A , which can result in reducing milk oxidative stability. On the other hand, the O R of mi lk D H A content and U l (%) on decreasing oxidative stability of mature human milk resulted in no significant effect. It is important to note that the tertile representing a high milk D H A content (tertile III, n=\3) ranged from 0.40-1.1 lg/ lOOg F A ; from which only six women expressed milk samples with D H A contents higher than 0.50g/100g F A , and from those six women only one resulted in a content higher than 0.80g/100g F A . Potential antioxidant effects of omega-3 L C - P U F A , such as D H A , as reported by Wander (2001) or any pro-oxidant effects due to greater relative oxidative decomposition (Banks, 1997; Turoli et al., 2004; Schneider, 2005) need to be studied more extensively. M i l k a-tocopherol was the only component that showed a statistically significant effect on the oxidative stability of mature human milk. The remaining milk variables resulted in a no real effect because 95% CI included ratios below and above 1.0 and thus functioned as both decreasing and increasing the odds of having low and/or high milk T A C values. The potential increasing effect on milk oxidative stability while presenting high contents of a-tocopherol (tertile III) ranged from 22 to >100%. This result strengthens the positive relationship found in this study between milk T A C and a-tocopherol content. However, the wide range of protection could also be due to a wide range in milk a-tocopherol contents in tertile III (2.71-5.02ug/mL). The study of a larger sample size would be important in further studies to confirm these findings. Despite the lack of a statistically significant effect, a 'pseudo' protective effect on milk oxidative stability may have also existed with the milk content of vitamin A (aM-trans-retinol) increasing from tertile I to III (OR 1.60 to 0.46) (Table 3.9). Higher T A C values 68 have been reported in infant formulas that contain higher vitamin A content (Turoli et al., 2004). A possible explanation why a high content of vitamin A (tertile III) did not increase the odds for having high milk T A C in the present study can be attributed to the actual low levels of vitamin A present in the tertile III in mature milk samples; however, more studies are needed to confirm these findings. 3.6 C O N C L U S I O N Participants in this particular study were characterized as being well educated women with relatively high household incomes. Patterns of maternal dietary intakes during gestation indicated healthy dietary habits and the common use of pre-natal multivitamin supplementation. Maternal dietary intakes were not used to evaluate potential relationships with human milk composition and T A C due to the difference in time between obtaining dietary intake data (16 and 36 weeks of gestation) and milk sampling (month 1 postnatally). The total antioxidant capacity (TAC) of 60 human milk samples was measured by the O R A C F L assay. Firstly, the oxidative stability of human milk was found to be attributed to milk a-tocopherol content. The positive relationship between this antioxidant vitamin and milk T A C was indeed significant and confirms the important role of vitamin E as a radical scavenger in milk, as previously reported for other biological systems. The lack of significant effects of other mi lk components analyzed in this study having, a role in modifying the oxidative stability in human milk could reflect the sum of antioxidant activity defined by other mechanisms of action not related to the peroxyl radical induced scavenging method examined herein. For example, non-enzymatic antioxidants such as citrate, phosphate, and sulphydryl groups, casein phosphopeptides and lactoferrin, and antioxidant enzymes could be included. The oxidative stability of human milk was not statistically affected by different contents of vitamin A (all-rrara'-retinol), polyunsaturated fatty acids, or by a calculated unsaturation index. Reported benefits of omega-3 fatty acids on antioxidant activity of biological systems also suggested the need of further research to substantiate a risk/benefit effect of elevated levels of L C - P U F A in milk. 69 GENERAL CONCLUSION The results of this research indicated that the total antioxidant capacity (TAC) of human milk against thermal-induced peroxyl radical attack can be measured by the widely used Oxygen Radical Absorbance Capacity assay, using fluorescein (FL) as the fluorescent probe. This methodology is known as O R A C F L - Advantages of O R A C F L assay include sensitivity, photostability, and high specificity. Standardization and validation experiments resulted in good linearity (R 2>0.990), precision (2.2%), and accuracy (>90%) of the assay. Determination of the optimal milk dilution factor was a very important factor for the accurate calculation of the O R A C F L value of milk samples with unknown antioxidant capacities. A dilution factor of 200x was chosen for routine O R A C F L analyses of human milk. In this study, a convenience sample consisting of 77 different human milk samples was obtained from different women at month 1 postnatally; from which 60 milk samples were tested for total antioxidant capacity (TAC) using the O R A C F L assay and analyzed for chemical composition (e.g. vitamin A and E, and fatty acids). Gestational dietary intakes were calculated from answered Food Frequency Questionnaire (FFQ) from a sub-sample of participants (o=29). Results indicated that gestational dietary intakes of these women were relatively stable and that pre-natal multivitamin supplementation was also common during gestation. Overall, despite the healthy dietary habits of all participants during gestation, it was also found that most women (82.8%) did not meet the recommendations for D H A through their diets. This result was consistent with previous studies reporting that lower intakes of D H A result in decreased human milk D H A concentrations. A limitation of this study was the absence of maternal dietary information during lactation, which corresponded to the actual time of human milk sampling. For this reason, it wi l l be important for future studies to relate maternal dietary intakes during lactation to mi lk antioxidant activity. Different compounds in human milk have been reported to possess antioxidant activity. Examples include the antioxidant vitamins and carotenoids, the peptides lactoferrin and glutathione, and the enzymes catalase, superoxide dismutase, and glutathione peroxidase. In this study, a potential antioxidant protection of human milk was identified to be attributed 70 to a-tocopherol, a vitamin E isomer and a natural chain-breaking antioxidant. Oxidative stability of mature human milk samples studied herein was significantly attributed to content of a-tocopherol. The lack of significant effects of other milk components on milk antioxidant activity also analyzed (e.g. vitamin A , polyunsaturated fatty acids, or unsaturation index), suggested that milk T A C could be affected by the sum of a complex action of different antioxidant mechanisms or other compounds, rather than by single components or specific mechanisms. According to the composition of mature human milk and the chemical and statistical analyses performed in this study, it was found that no significant risk for decreasing the oxidative stability of human milk existed with the different components examined herein. In fact, it appears that the common result of consuming dietary pre-natal multivitamin supplements during pregnancy and lactation may have been a factor in explaining the relative high relationship between milk vitamin E content and milk T A C , and between vitamin E and vitamin A contents in mature human milk. The observation that milk P U F A and L C - P U F A contents, in particular omega-3 L C - P U F A , as well as unsaturation index, were not factors in predisposing mother's milk to change in oxidative stability could be re-evaluated closely using a greater range of minimum to maximum levels than those obtained herein in order to determine any potential antioxidant or pro-oxidant effect in human milk, as has been previously found in other biological systems. In general, it can be concluded that human milk represents a source of potential beneficial components for the breastfed neonate. 71 REFERENCES Agostoni, C. 2005. L C - P U F A Content in Human M i l k : Is It A lways Optimal? Acta Paediatrica, 94: 1532- 1537. A l , M .D , v. Houwelingen, A . C , Badart-Smook, A . , and Hornstra, G . 1995. Some Aspects of Neonatal Essential Fatty A c i d Status are altered by Linoleic Ac i d Supplementation of Women during Pregnancy. Journal of Nutrition, 725: 2822-2830. Alberti-Fidanza, A . , Bur ini , G. , arid Perriello, G. 2002. Total Antioxidant Capacity of Colostrum, and Transitional and Mature Human Mi lk . The Journal of Maternal-Fetal and Neonatal Medicine, IP. 275-279. A l i , J . , Kader, A . , Hassan, K., and Arshat, H. 1986. Changes in Human M i l k Vitamin E and Total Lipids during the First Twelve Days of Lactation. The American Journal of Clinical Nutrition, 43: 925-930. A l len, L. H. 2005. Mult iple Micronutrients in Pregnancy and Lactation: A n Overview. American Journal of Clinical Nutrition, 8 P. 1206S-1212S. Aycicek, A . , Erel , O., Kocyigi t , A . , Selek, S., and Demirkol, M.R. 2006. Breast M i l k Provides Better Antioxidant Power than does Formula. Nutrition, 22: 616-619. Balk, E. M . , Lichtenstein, A . H., Chung, M . , Kupelnick, B., Chew, P., and Lau, P. 2006. Effects on Omega-3 Fatty Acids on Serum Markers of Cardiovascular Disease Risk: A Systematic Review. Atherosclerosis, 189: 19-30. Banks, M . A . 1997. Chapter 4: Vi tamin E and Polyunsaturated Fats: Antioxidant and Pro-oxidant Relationship. In: Oxidants, Antioxidants and Free Radicals. Edited by: Steve Baskin and Harry Salem. Taylor and Francis Publishers. Pp. 95-112. 72 Barbosa, D. S., Cecchini, M . Z., E l Kadri , M . Z. , Rodriguez, M . A . M . , Burini , R. C , and Dichi , I. 2003. Decreased Oxidative Stress in Patients with Ulcerative Colit is Supplemented With Fish O i l co-3 Fatty Acids. Nutrition, 19: 837-842. Baydas, G . , Karatas, F., Gursu, M . F., Bozkurt, H. A . , Ilhan, N . , Yasar, A . , and Canatan, H. 2002. Antioxidant Vitamins Levels in Term and Preterm Infants and their Relation to Maternal Vi tamin Status. Archives of Medical Research, 33: 276-280. Boersma, E. R., Offringa, P. J . , Muskiet, F .A.J . Chase, W . M . and Simmons, I.J. 1991. Vitamin E, L ip id Fractions, and Fatty Ac id Composition of Colostrum, Transitional Mi lk , and Mature M i l k : A n International Comparative Study. American Journal of Clinical Nutriton, 53: 1197'-1204. Bohles, H. 1997. Antioxidative Vitamins in Prematurely and Maturely Born Infants. International Journal for Vitamin and Nutrition Research, 67: 321-328. Bondet, V . , Brand-Wil l iams, W., and Berset, C. 1997. Kinetics and Mechanisms of Antioxidant Act iv i ty using the D P P H Free Radical Method. Food Science & Technology, 30: 609-615. Borna, S., Borna, H . and Daneshbodie, B. 2005. Vitamins C and E in the Latency Period in Women with Preterm Premature Rupture of Membranes. International Journal of Gynecology and Obstetrics, 90: 16-20. Bourre, J - M . 2007. Dietary Omega-3 Fatty Acids for Women. Biomedicine & Pharmacotherapy: 1-8 Bramley, P .M . , Elmadfa, I., Kafatos, A . , Kel ly , F.J., Manios, Y . , Roxborough, H.E., Schuch, W., Sheehy, P.J.A., and Wagner, K . H . 2000. Vitamin E. Journal of the Science of Food and Agriculture, 80: 913-938. 73 Cadenas, E. 1998. Chapter 22: Mechanisms of Antioxidant Act ion. In: Free Radicals, Oxidative Stress, and Antioxidants. Edited by: Tomris Ozben. N e w York. Plenum Press. Pp. 237-252. Canfield, L. M . , Kamisnky, R. G. , Taren, D. L., Shaw, E., and Sander, J . K. 2001. Red Palm O i l in the Maternal Diet Increases Provitamin A Carotenoids in Breastmilk and Serum of the Mother-Infant Dyad. European Journal of Nutrition, 40: 30-38. Cao, G. , Alessio, H. M . , and Cutler, R. G. 1993. Oxygen Radical Absorbance Capacity Assay for Antioxidants. Free Radical Biology & Medicine, 14: 303-311. Cao, G. , Booth, S. L., Sadowski, J . A . and Prior, R. L. 1998. Increases in human plasma antioxidant capacity after consumption of controlled diets high in fruit and vegetables. American Journal of Clinical Nutrition, 68: 1081-1087. Cao, G. and Prior, R. L. 1999. The Measurement of Oxygen Radical Absorbance Capacity in Biological Samples. Methods in Enzymology, 299: 50-62. Cao, G. and Prior, R. L. 2002. Ch.2: Measurement of Total Antioxidant Capacity in Nutritional and Cl in ical Studies. In: Handbook of Antioxidants. Ed . Enrique Cardenas and Lester Packer. Marcel Dekker. U S A . Pp. 47-56. Cano, A . , Acosta, M . , and Arnao, M .B . 2003. Hydrophil ic and Lipophil ic Antioxidant Activity Changes during On-Vine Ripening of Tomatoes (Lycopersicon esculentum Mill.). Postharvest Biology and Technology, 28: 59-65. Carlson, S.E. , Cooke, R.J . , Werkman, S.H. , and Tolley, E A . 1992. First Year Growth of Preterm Infants Fed Standard compared to Marine O i l n-3 Supplemented Formula. Lipids, 27: 901-907. 74 Casanueva, E., Magana, L., Pfeffer, F. And Baez, A . 1991. Incidence of Premature Rupture of Membranes in Pregnant Women with Low Leukocyte Levels of Vi tamin C. European Journal of Clinical Nutrition, 45: 401-405. Croft, K. D. 1998. The Chemistry and Biological Effects of Flavonoids and Phenolic Acids Annals of the New York Academy of Sciences 854, 435-442. Daneel-Otterbech, S., Davidsson, L. and Hurrell, R. 2005. Ascorbic A c i d Supplementation and Regular Consumption of Fresh Orange Juice Increase the Ascorbic Ac i d Content of Human M i l k : studies in European and African Lactating Women. American Journal of clinical Nutrition, 8E. 1088-1093. Davalos, A . , Gomez-Cordoves, C , and Bartolome, B. 2004. Extending Applicabil i ty of the Oxygen Radical Absorbance Capacity (ORAC-Fluorescein) Assay. Journal of Agriculture and Food Chemistry, 52: 48-54. Davies, M . J . , Fu , S., Wang, H., and Dean, R. T. 1999. Stable Markers of Oxidant Damage to Proteins and their Application in the Study of Human Disease. Free Radical Biology & Medicine, 27: 1151-1163. Denomme, J . , Stark, K. D., and Holub, B. J . 2005. Directly Quantitated Dietary (n-3) Fatty Ac i d Intakes of Pregnant Canadian Women are Lower than Current Dietary Recommendations. Journal of Nutrition, 755:206-211. Elias, S. L. and Innis, S. M . 2001. Infant Plasma trans, n-6, and n-3 Fatty Acids and conjugated Linoleic Ac ids are Related to Maternal Plasma Fatty Acids, Length of Gestation, and Birth Weight and Lentght. American Journal of Clinical Nutrition, 73: 807-814. El is ia, I., Hu , C , Popovich, D. G . , and Kitts, D. D. 2007. Antioxidant Assessment of an Anthocyanin-enriched Blackberry Extract. Food Chemistry, 10E. 1052-1058. 75 Engler, M . B . and Engler, M . M . 2006. The Emerging Role of Flavonoid-Rich Cocoa and Chocolate in Cardiovascular Health and Disease. Nutrition Reviews, 64: 109-118. Erdman, J .W, Balentine, D., Arab, L., Beecher, G . , Dwyer, J . T., Folts, J . , Harnly, J . , Hollman, P., Keen, C .L . , Mazza, G . , Messine, M . , Scalbert, A . , V i ta , J . , Wil l iamson, G. , and Burrowes, J . 2007. Flavonoids and Heart Health: Proceedings of the ILSI North America Flavonoids Workshop. Journal of Nutrition, 137: 718S-737S. E S H A Food Processor Manual. Appendix A-14: Conversions Vi tamin A and Appendix A -15: Conversions Vi tamin E. Access: May 2007 < http://www.esha.com/support/fp> Ettinger, S. 2004. Macronutrients: Carbohydrates, Proteins, and Lipids. In: Krause's Food, Nutrition & Diet Therapy. 11 t h Ed. U S A . Pp. 37-74. Fernandez-Pachon, M . S., Vi l lano, D., Troncoso, A . M . , and Garcia-Parril la, M . C. 2005. Antioxidant Capacity of Plasma after Red Wine Intake in Human Volunteers. Journal of Agricultural and Food Chemistry, 53: 5024-5029. Francois, C .A . , Connor, S.L., Wander, R.C. , and Conner, W .E . 1998. Acute Effects of Dietary Fatty Ac ids on the Fatty Acids of Human Mi lk . American Journal of Clinical Nutrition, 67: 301-308. Friel, J.K., Mart in, S. M . , Langdon, M . , Herzberg, G.R., and Buettner, G.R. 2002. M i l k from Mothers of Both Premature and Full-Term Infants Provides Better Antioxidant Protection than Does Infant Formula. Pediatric Research, 51: 612-618. Gallagher, M . L. 2004. Vitamins. In: Krause's Food, Nutrition & Diet Therapy. 11 t h Ed. U S A . Pp. 76, 88, 110. 76 Garb, J . L. 1996. Understanding Medical Research: A Practitioner's Guide. L, B and Co, Inc. U S A . Pp. 212-213. Gibson, R A . and Makrides, M . 1998. The Role of Long Chain Polyunsaturated Fatty Acids ( L C P U F A ) in Neonatal Nutrition. Acta Paediatrica, 87: 1017-1022. Gopal, P. K. and G i l l , H. S. 2000. Oligosaccharides and Glycoconjugates in Bovine M i l k and Colostrum. British Journal of Nutrition, 84: S69 - S74. Gossage, C P , Deyhim, M . , Yamimi , S., Douglass, L.W., and Moser-Vei l lon, P.B. 2002. Carotenoid Composition of Human M i l k during the First Month Postpartum and the Response to P-carotene Supplementation. American Journal of Clinical Nutrition, 76: 193-197. Granot, E., Golan, D., R ivk in , L., and Kohen R. 1999. Oxidatives stress in healthy breast fed versus formula fed infants. Nutrition Research, 19: 869-879. Gutierrez, M . , Sola, M . M . , and Vargas, A . M . 2005. Fatty A c i d Composition of Phospholipids in Mesocarp of Cherimoya Fruit during Ripening. Food Chemistry 90: 341— 346. Hadders-Algra, M . , Bouwstra, H., Goor, S. A . , Dijck-Brouwer, D. A . J . , and Muskiet, F. A . J . 2007. Prenatal and Early Postnatal Fatty Ac id Status and Neurodevelopmental Outcome. Journal of Perinatal Medicine, 35: S28-S34. Hal l iwel l , B. and Gutteridge, J . M . C. 1999. Free Radicals in Biology and Medicine. New York. Oxford University Press. Pp. 36-100. Hamosh, M . 2001. Bioactive Factors in Human Mi lk . Pediatric Clinics of North America, 48: 69-86. 77 Hanna, N . , Ahmed, K., Anwar, M . , Petrova, A . , Hiatt, M . , and Hegyi , T. 2004. Effect of Storage on Breast M i l k Antioxidant Activity. Archives of Disease in Childhood: Fetal and Neonatal Edition, 89: F518-F520. Health Canada. 1999. Nutrition for a Healthy Pregnancy: National Guides for the Childbearing Years. Minister of Public Works and Government Services Canada. Ottawa. Health Canada. Dietary Reference Intakes: Unit Conversion Factors. Access: May, 2007. <http://www.hc-sc.gc.ca/fn-an/nutrition/reference/table/index_e.html> Health Canada. Drug and Health Products Monographs. Conversion Factors for Supplements. Access: May , 2007. <http://www.hc-sc.gc.ca/dhp-mps/prodnatur/applications/ licen-prod/monograph/mono_vitamin_a_e.html> Heim, K. E., Tagliaferro, A . R., Bobi lya, D. J . 2002. Flavonoid Antioxidants: Chemistry, Metabolism and Structure-Activity Relationships. Journal of Nutritional Biochemistry, 13: 572-584. Helland, I. B., Saugstad, O. D., Saarem, K., Houwelingen, A . C , Nylander, G. , and Drevon, C. A . 2006. Supplementation of N-3 Fatty Acids during Pregnancy and Lactation Reduced Maternal Plasma L ip id Levels and Provides D H A to the Infants. The Journal of Maternal-Fetal and Neonatal Medicine, 19: 397-406. Helzlsouer, K .J . , Huang, H.Y. , Alberg, A . J . , Hoffman, S., Burke, A . , Norkus, E.P., Morris, J.S., Comstock, G.W. 2000. Association between a-Tocopherol, y-Tocopherol, Selenium, and Subsequent Prostate Cancer. Journal of the National Cancer Institute, 92: 2018-2023. Hu , C. and Kitts, D. D. 2001. Evaluation of Antioxidant Activi ty of Epigallocatechin Gallate in Biphasic Model Systems in vitro. Molecular and Cellular Biology, 218: 147-155. 78 Hu, C , Kwok, B. H. L., and Kitts, D. D. 2005. Saskatoon Berries (Amelanchier alnifolia Nutt.) Scavenge Free Radicals and Inhibit Intracellular Oxidation. Food Research International, 38: 1079-1085. Huang, D., Ou, B., and Prior, R. L. 2005. The Chemistry behind Antioxidant Capacity Assays. Journal of Agricultural and Food Chemistry, 53: 1841-1856. Hwang, E-S. and K i m , G - H . 2007. Biomarkers for Oxidative Stress Status of D N A , lipids, and Proteins in vitro and in vivo Cancer Research. Toxicology, 229: 1-10. Innis, S . M . 2003. Perinatal Biochemistry and Physiology of Long-Chain Polyunsaturated Fatty Acids. Journal of Biochemistry, 143: S1-S8. Innis, S . M . and El ias, S. L. 2003. Intakes of Essential n-6 and n-3 Polyunsaturated Fatty Acids among Pregnant Canadian Women. American Journal of Clinical Nutrition, 77: 473-478. Innis, S .M . , Gi l ley, J . , and Werker, J . 2001. Are Human M i l k Long-Chain Polyunsaturated Fatty Acids Related to Visual and Neural Development in Breast-fed Term Infants? The Jounal of Pediatrics, 139: 532-538. Institute of Medicine. 2000. Dietary Reference Intakes for Vi tamin A , Vitamin K, Arsenic, Boron, Chromium, Copper, Iodine, Iron, Manganese, Molybdenum, Nickel , Sil icon, Vanadium, and Zinc. National Academy Press, U S A . <http://books.nap.edu/openbook.php?record_id=10026&page=82> Jacob, R. A . 1995. The Integrated Antioxidant System. Nutrition Research, 15(5): 755-766. Jackson, M . J . , Khassaf, M . , Vasi laki , A . , McArd le , F., and McArd le , A . 2004. Vitamin E and the Oxidative Stress of Exercise. Annals of the New York Academy of Sciences, 1031: 158-168. 79 Jansson, L., Akesson, B., Holmberg, L. 1981. Vitamin E and Fatty Ac i d Composition of Human Mi lk . The American Journal of Clinical Nutrition, 34: 8-13. Jensen, C.L . , Maude, M . , Anderson, R.E. , and Heird, W . C . 2000. Effect of Docosahexaenoic Ac i d Supplementation of Lactating Women on the Fatty Ac i d Composition of Breast M i l k Lipids and Maternal and Infant Plasma Phospholipids. American Journal of Clinical Nutrition, 71: 292S-299S. Jian, Q., Christen, S., Shigenaga, M.K. , and Ames, B .N . 2001. y-Tocopherol, the major form of Vitamin E in the U S Diet, Deserves More Attention. American Journal of Clinical Nutrition, 74: 714-722. Kasapovic, J . , Pejic, S., Mladenovic, M . , Radlovic, N . , and Pajovic, S. B. 2005. Superoxide Dismutase Act ivi ty in Colostrum, Transitional and Mature Human M i l k . The Turkish Journal of Pediatrics, 47:1>A2>-1>A1. Kitts, D.D. 1993. Bioactive Substances in Food: Identification and Potential Uses. Canadian Journal of Physiology and Pharmacology, 72: 423-434. Kitts, D.D. 1997. A n Evaluation of the Multiple Effects of the Antioxidant Vitamins. Trends in Food Science and Technology, 8: 198-203. Kitts, D.D. 2005. Antioxidant Properties of Casein-Phosphopeptides. Trends in Food Science & Technology, 16: 549-554. Kitts, D.D. 2006. Chapter 3: Antioxidant Phytochemicals and Potential Synergistic Activities at Reducing Risk of Cardiovascular Disease. In: Food-Drug Synergy and Safety. Edited by L i l ian V . Thompson and Wendy E. Ward. Taylor & Francis Group. U S A . p.27. 80 Kitts, D.D., Wijewickreme, A . N . , and Hu, C. 2000. Antioxidant Properties of a North American Ginseng Extract. Molecular and Cellular Biochemistry, 203: 1-10. Kitts, D.D. and Hu , C. 2005. Biological and Chemical Assessment of Antioxidant Activity of Sugar-Lysine Model Mai l lard Reaction Products. Annals of the New York Academy of Sciences, 1043: 501-512. Korchazhkina, O., Jones, E., Czauderna, M . , Spencer, S. A . , and Kowalczyk, J . 2006. H P L C with U V Detection for Measurement of Vitamin E in Human M i l k . Acta Chromatographica, 16: 48-57. Kurlak, L. O. and Stephenson, T. J . 1999. Plausible Explanations for Effects of Long Chain Polyunsaturated Fatty Ac ids ( L C - P U F A ) on Neonates. Archives of Disease in Childhood: Fetal and Neonatal Edition, 80: F148-F154. L 'Abbe, M.R. and Friel , J .K. 2000. Superoxide Dismutase and Glutathione Peroxidase Content of Human M i l k From Mothers of Premature and Ful l-Term Infants During the First 3 Months of Lactation. Journal of Pediatric Gastroenterology and Nutrition, 57:270-274. Ladipo, O. A . 2000. Nutrit ion in Pregnancy: Mineral and Vitamin Supplements. American Journal of Clinical Nutrition, 72: 280S-290S. Lawrence, R.A. 1994. Chapter 4: Biochemistry of Human Mi lk . In: Breastfeeding. 4 t h Ed. U S A . Pp. 91-148. Leonard, S.W., Paterson, E., Atkinson, J.K., Ramakrishnan, R., Cross, C E . and Traber, M . G . 2005. Studies in Humans using Deuterium-labeled a- and y-Tocopherols Demonstrate Faster Plasma y-tocopherol Disappearance and Greater y-Metabolite Production. Free Radical Biology and Medicine, 38: 857-866 81 LePage, G . and Roy, C. C. 1986. Direct transesterification of all classes of lipids in a one-step reaction. Journal of Lipid Research, 27: 114-120. L i m Y . Y . , L i m , T. T., and Tee, J . J . 2007. Antioxidant Properties of Several Tropical Fruits: A Comparative Study. Food Chemistry, 103: 1003-1008. Lindmark-Mansson, H. and Akesson, B. 2000. Antioxidative Factors in Mi l k . British Journal of Nutrition, 84: S103-S110. L iu , R. H. 2003. Health Benefits of Fruits and Vegetables are from Addit ive and Synergistic Combinations of Phytochemicals. American Journal of Clinical Nutrition, 78: 517S-520S. Livrea, M . A . and Tesoriere, L. 1998. Ch. 5: Antioxidant Act iv i ty o f Vitamin A within L ip id Environments. In: Subcellular Biochemistry: Fat Soluble Vitamins. Ed. Quinn, P. J. and Kagan, V . E. Springer. Pp. 115-144. Macias, C. and Schweigert, F.J. 2001. Changes in the Concentration of Carotenoids, Vitamin A , Alpha-Tocopherol and Total Lipids in Human M i l k thruought Early Lactation. Annals of Nutrition and Metabolism, 45: 82-85. Malacarne, M . , Martuzzi , F., Summer, A . , and Mariani, P. 2002. Protein and Fat Composition of Mare's M i l k : Some Nutritional Remarks with Reference to Human and Cow's M i l k . International Dairy Journal, 72:869-877. Mastaloudis, A . , Leonard, S. W., and Traber, M . G . 2001. Oxidative Stress in Athletes during Extreme Endurance Exercise. Free Radical Biology & Medicine, 57(7): 911-922. Matsubasa, T., Uchino, T., Karashima, S., Tanimura, M . , and Endo, F. 2002. Oxidative Stress in Very L o w Birth Weight Infants as Measured by Urinary 8-OHdG. Free Radical Research, 2: 189-193. 82 Meneses, F. and Trugo, N . M . F. 2005. Retinol, P-carotene, and Lutein + Zeaxanthin in the M i l k of Brazi l ian Nursing Women: Associations with Plasma Concentrations and Influences of Maternal Characteristics. Nutrition Research, 25: 443-451. Mermen, L. I., Saphino, D., Ito, H., Bertrais, S., Galan, P., Hercberg, S., and Scalbert, A . 2006. Urinary Flavonoids and Phenolic Acids as Biomarkers of Intake for Polyphenol-Rich Foods. British Journal of Nutrition, 96: 191-198. Mil iauskas, G . , Venskutonis, P. R., van Beek, T. A . 2003. Screening of radical scavenging activity of some medicinal and aromatic plant extracts. Food Chemistry, 85: 231-237. Miquel , E., Alegria, A . , Barbera, R., Farre, R., and Clemente, G . 2004. Stability of Tocopherols in Adapted Mi lk-Based Infant Formulas during Storage. International Dairy Journal, 14: 1003-1011. Miranda, M . , Muriach, M . , Almansa, I., and Jareno, E. 2004. Oxidative Status of Human M i l k and its Variations during Cold Storage. Biofactors, 20: 129-137. Mor i , T. A . 2004. Effect of Fish and Fish Oil-derived Omega-3 Fatty Acids on Lip id Oxidation. Redox Report, 9: 193-197. Morris, J . M . , Gopaul, N . K., Endresen, M . J . R., Knight, M . , Linton, E. A . , Dhir, S., Anggard, E. E., and Redman, C. W. G. 1998. Circulating Markers of Oxidative Stress are Raised in Normal Pregnancy and Pre-Eclampsia. British Journal of Obstetrics and Gynaecology, 105: 1195-1199. Morrow, A . L., Ruiz-Palacios, G . M . , Altaye, M . , Jiang, X . , Guerrero, M . L., Meinzen-Derr, J . K., Farkas, T., Chaturvedi, P., Pickering, L. K., Newburg, D. S. 2004. Human M i l k Oligosaccharides are Associated with Protection against Diarrhea in Breast-fed Infants. Journal of Pediatrics, 145: 297-303 83 Newton, E. R. 2004. Breastmilk: the Gold Standard. Clinical Obstetrics and Gynecology, 47: 632-642. Nourooz-Zadeh, J . , Tajaddini-Sarmadi, J . , L ing, K .L .E . , and Wolff , S.P. 1996. Low-Density Lipoprotein is the Major Carrier of L ip id Hydroperoxides in Plasma. Biochemical Journal, 313: 781-786 Olafsdottir, A . S., Wagner, K - H . , Thorsdottir, I. 2001. Fat-Soluble Vitamins in the Maternal Diet, Influence of Cod Liver O i l Supplementation and Impact of the Maternal Diet on Human M i l k Composition. Annals of Nutrition and Metabolism, 45: 265-272. Ortega, R. M . , Andres, P., Martinez, R. M . , and Lopez-Sobaler, A . M . 1997. Vitamin A Status during the Third Trimester of Pregnancy in Spanish Women: Influence on Concentration of Vi tamin A in Breast Mi lk . American Journal of Clinical Nutrition, 66: 564-568. Ortega, R. M . , Quintas, M . E., Andres, P., Martinez, R. M . , Lopez-Sobaler, A . M . 1998. Ascorbic A c i d Levels in Maternal M i l k : Differences with respect to Ascorbic Ac id Status during the Third Trimester of Pregnancy. British Journal of Nutrition, 79: 431-437. Ortega, R. M . , Lopez-Sobaler, A . M . , Andres, P., Martinez, R. M . , Quintas, M . E., and Requejo, A . M . 1999. Maternal Vitamin E Status during the Third Trimester of Pregnancy in Spanish Women: Influence on Breast M i l k Vitamin E Concentration. Nutrition Research, 19: 25-36. Oostenbrug, G . S., Mensink, R. P., A l , M . D. M . , Houwelingen, A . C , and Hornstra, G. 1998. Maternal and Neonatal Plasma Antioxidants Levels in Normal Pregnancy, and the Relationship with Fatty Ac i d Unsaturation. British Journal of Nutrition, 80: 61-13. 84 Ou, B., Hampsch-Woodi l l , M . , and Prior, R. L. 2001. Development and Validation of an Improved Oxygen Radical Absorbance Capacity Assay using Fluorescein as the Fluorescent Probe. Journal of Agricultural and Food Chemistry, 49: 4619-4626. Phipps, S .M . , Sharaf, M . H . M . , and Butterweck, V . 2007. Antioxidant Activi ty in Botanicals and Other Dietary Supplements. Pharmacopeial Forum, 33(4) July-August 2007. Picciano, M . F. 1998. Human M i l k : Nutritional Aspects of a Dynamic Food. Biology of Neonate, 74: 84-93. Picciano, M . F. 2001. Nutrient Composition of Human M i l k . Pediatric Clinics of North America, 48: 53. Picciano, M . F. 2003. Pregnancy and Lactation: Physionological Adjustments, Nutritional Requirements and the Role of Dietary Supplements. Journal of Nutrition, 133: 1997-2002S. Portney, L .G . and Watkins, M.P. 2000. Foundations of Cl in ical Research: Applications to Practice. 2 n d Ed . U S A . Pp. 382, 674-677. Pressman, E. K., Cavanaugh, J . L., Mingione, M . , Norkus, E. P., and Woods, J . R. 2003. Effects of Maternal Antioxidant Supplementation on Maternal and Fetal Antioxidant Levels: A Randomized, Double-blind Study. American Journal of Obstetrics and Gynecology, 189, 1720-1725. Prior, R. L. 2004. Biochemical Measures of Antioxidant Status. Topics in Clinical Nutrition, 19 (3): 226-238. Prior, R. L. and Cao, G . 1999. In V i vo Total Antioxidant Capacity: Comparison of Different Analytical Methods. Free Radical Biology and Medicine, 27(\ 1/12): 1173-1181. 85 Prior, R L . , Hoang, H., Gu , L., Wu , X . , et al. 2003. Assays for hydrophilic and lipophilic antioxidant capacity (oxygen radical absorbance capacity ( O R A C F L ) of plasma and other biological and food samples. Journal of Agricultural and Food Chemistry, 51: 3273-3279. Ramirez, M . , Gallardo, E . M . , Souto, A . S . , Weissheimer, C , and G i l , A . 2001. Plasma Fatty-Ac i d Composition and Antioxidant Capacity in Low Birth-Weight Infants Fed Formula Enriched with n-6 and n-3 Long-Chain Polyunsaturated Fatty Acids from Purified Phospholipids. Clinical Nutrition, 20: 69-76 Ribeiro, K. D. S., Me lo , I. L. P., Pristo, A . Z. O., and Dimenstein, R. 2005. The Effect of Processing on the Vi tamin A content of Human Mi lk . Journal de Pediatria, 81: 61-64. Rietjens, I., Boersma, M . G. , de Haan, L., Spenkelink, B., Awad, H . M . , Cnubben, N. , van Zanden, J . J . , van der Woude, H., A l ink, G. M . , and Koeman, J . H. 2002. The pro-oxidant Chemistry of the Natural Antioxidants Vitamin C, Vi tamin E, Carotenoids and Flavonoids. Environmental Toxicology and Pharmacology, 11: 321-333. Rodas Mendoza, B., Morera Pons, S., Castellote Bargallo, A . I., and Lopez-Sabater, M . C. 2003. Rapid Determination by Reverse-Phase High Performance L iquid chromatography of Vitamins A and E in Infant Formulas. Journal of Chromatography, 1018: 197-202. Salem, H. and Baskin, S. I. 1997. Chapter 13: Toxicity of Oxygen and Ozone. In: Oxidants, Antioxidants and Free Radicals. Edited by: Steve Baskin and Harry Salem. United Kingdon. Taylor and Francis Publishers. Pp.228. Schacky, C. and Harris, W . S. 2007. Cardiovascular Benefits of Omega-3 Fatty Acids. Cardiovascular Research, 73: 310-315. Schlesier, K., Harwat, M . , Bohm, V . , and Bitsch, R. 2002. Assessment of Antioxidant Activity by using Different in vitro Methods. Free Radical Research, 36: 177-187. 86 Schneider, C. 2005. Chemistry and Biology of Vitamin E. Molecular Nutrition and Food Research, 49: 7-30. Schweigert, F.J. , Hurtienne, A . , and Bathe, K. 2000. International Journal for Vitamin and Nutrition Research, 70:79-83. Semba, R.D. 1999. Vi tamin A as "Anti-Infective" Therapy, 1920-1940. Journal of Nutrition, 129: 783-791. Shahidi, F. 2004. Functional Foods: Their Role in Health Promotion and Disease Prevention. Journal of Food Science, 69: R146-R149. Shoji, H., Shinohara, K., Oguchi, S., Shiga, S., and Yamashiro, Y . 2004. Suppressive Effects of Breast M i l k on Oxidative D N A Damage in Very L o w Birthweight Infants. Archives of Disease in Childhood: Fetal and Neonatal Edition, 89: F136-F138. Shoji, H., Oguchi, S., Fujinaga, S., Shinohara, K., Kaneko, K., Shimizu, T., and Yamashiro, Y . 2005. Effects of Human M i l k and Spermine on Hydrogen Peroxide-Induced Oxidative Damage in IEC-6 Cells. Journal of Pediatric Gastroenterology and Nutrition, 41: 460-465. Si lva, M .H .L , Si lva, M .T .C . , Brandao, S.C.C. , Gomes, J .C. , Peternelli, L A . , Franceschini, S. do C C . 2005. Fatty A c i d Composition of Mature Breast M i l k in Brazil ian Women. Food Chemistry, 93: 297- 303. Simopoulos, A . P . , Leaf, A . , and Salem, N . 1999. Workshop on the Essentiality of and Recommended Dietary Intakes for Omega-6 and Omega-3 Fatty Acids. Journal of the American College of Nutrition, 18: 487-489. Sokol, R. J . 1988. Vi tamin E Deficiency and Neurologic Disease. Annual Review of Nutrition, 8: 351-373. 87 Temay, A . L. and Sorokin, V . 1997. Chapter 1: Redox, Radicals, and Antioxidants. In: Oxidants, Antioxidants and Free Radicals. Edited by: Steve Baskin and Harry Salem. United Kingdon. Taylor and Francis Publishers. Pp. 1-16. Tesoriere, L., Bongiorno, A . , Pintaudi, A . M . , D 'Anna, R., D 'Arpa , D., and Livrea, M . A . 1996. Synergistic Interactions between Vitamin A and Vitamin E against L ip id Peroxidation in Phosphatidylcholine Liposomes. Archives of Biochemistry and Biophysics. 326: 57-63. Therond, P., Bonnefont-Rousselot, D., Davit-Spraul, A . , Conti, M . , and Legrand, A . 2000. Biomarkers of oxidative stress: an analytical approach. Current Opinion in Clinical Nutrition and Metabolic Care, 3: 373-384. Thompson, F. E. and Byers, T. 1994. Dietary Assessment Resource Manual. Journal of Nutrition, 124: 2245S-2317S. Traber, M . G . 1999. Uti l ization of Vitamin E. Biofactors, 10: 115-120. Tsuzuki, W., Yunok i , R., and Yoshimura, H. 2007. Intestinal Epithelial Cells Absorb y-Tocotrienol Faster than a-Tocopherol. Lipids, 42: 163-170. Turoli , D., Testolin, G. , Zanini, R., and Bel lu, R. 2004. Determination of Oxidative Status in Breast and Formula M i l k . Acta Paediatrica, 93: 1569-1574. Ubaldi , A . , Delbono, G . , Fusari, A . and Serventi, P. 2005. Quick H P L C Method to Determine Vi tamin E Concentration in Cow's Mi lk . Ann. Fac. Medic. Vet. di Parma, 25: 101-110. Urso, M . L. and Clarkson, P. M . 2003. Oxidative Stress, Exercise, and Antioxidant Supplementation. Toxicology, 189: 41-54. 88 U S D A Agriculture Research Service. 2002. Appendix B: Food Sources of Vitamin A , C and E. Nutrient Database for Standard Reference, Release 17. U S D A National Nutrient Database for Standard Reference, Release 20. 2007. <http://wvvw.nal.usda.gov/fnic/foodcomp/cgi-bin/nut_search_new.pl> VanderJagt, D.J., Ojolo, S .N. , Constanza, A . , Blackwel l , W. , and Glew, R .H. 2001. Antioxidant Content of the M i l k of Nigerian Women and the Sera o f their Exclusively Breast-fed Infants. Nutrition Research, 21: 121-128. V is io l i , F., Colombo, C , and Gal l i , C. 1998. Oxidation of Individual Fatty Acids Yields Different Profiles of Oxidation Markers. Biomedical and Biophysical Research Communications, 245: 487-489. Wander, R .C. 2001. Ch . 19: L ip id Oxidation in Biological Systems Enriched with Long Chain n-3 Fatty Acids. In: Handbook of Nutraceuticals and Functional Foods. C R C Press. U S A . Pp. 305-330. Watkins, B. A . , Hannon, K., Ferruzi, M . , L i , Y . 2007. Dietary P U F A and Flavonoids as Deterrents for Environmental Pollutants. Journal of Nutritional Biochemistry, 18: 196-205. Whitley, E. and Ba l l , J . 2002. Statistics review 3: Hypothesis testing and P values. Critical Care, 6: 222-225. Wong, P .Y .Y . , Nakamura, S., and Kitts, D.D. 2006. Functional and Biological Activities of Casein Glycomacropeptide as Influenced by Lipophil ization with Medium and Long Chain Fatty Ac id . Food Chemistry, 97: 310-317 Wood, R.D., Mi tchel l , M . Sgouros, J .G. , and Lindahl, T. 2001. Human D N A Repair Genes. Science 291: 1284-1289. 89 Worthington-Roberts, B. S. 1997. Chapter 12: Human M i l k Composition and Infants Growth and Development. In: Nutrition in Pregnancy and Lactation. Edited by: Bonnie S. Worthington-Roberts and Sue Rodwell Wil l iams. Brown & Benchmark Publishers. U S A . Pp. 345-384. Wu , X . , Gu , L., Holden, J . , Haytowitz, D. B., Gebhardt, S. E., Beecher, G . , and Prior, R. L. 2004. Development of a Database for Total Antioxidant Capacity in Foods: a Preliminary Study. Journal of Food Composition and Analysis, 17: 407-422. Xiang, M . , Harbige, L. S., and Zetterstrom, R. 2005. Long-chain Polyunsaturated Fatty Acids in Chinese and Swedish Mothers: Diet, Breast M i l k and Infant Growth. Acta Paediatrica, 94: 1543-1549. X u , Z. , Hua, N . , Godber, J.S. 2001. Antioxidant Activity of Tocopherols, Tocotrienols, and y-Oryzanol Components from Rice Bran Against Cholesterol Oxidation Accelerated by 2,2'-azobis (2-methylpropionamidine) dihydrochloride. Journal of Agricultural and Food Chemistry, 49: 2077-2781. Yeum, K., Ferlan, G . , Patty, J . , Russell, R. M . 1998. Relationship of Plasma Carotenoids, Retinol and Tocopherols in Mothers and Newborn Infants. Journal of the American College of Nutrition, 17: 442-447. Yeum, K., Russell , R. M . , Kr insky, N . I., and A ld in i , G . 2004. Biomarkers of Antioxidant Capacity in the Hydrophil ic and Lipophil ic Compartments of Human Plasma. Archives of Biochemistry and Biophysics, 430: 97-103. Yuan, Y . V . and Kitts, D.D. 1997. Chapter 15: Endogenous Antioxidants: Role of Antioxidant Enzymes in Biological Systems. In: Natural Antioxidants Chemistry, Health Effects, and Applications. Edited by Fereidoon Shahidi. A O C S Press. U S A . Pp. 258-270. 90 Zoeren-Grobben, D., Lindeman, J . H. N. , Houdkamp, E., Brand, R., Schrijver, J . , and Berger, H. M . 1994. Postnatal Changes in Plasma Chain-Breaking Antioxidants in Healthy Preterm Infants Fed Formula and/or Human Mi lk . American Journal of Clinical Nutrition, 60: 900-906. 91 APPENDIX A: MATERNAL DIETARY INTAKES A.l Maternal Dietary Intakes Answered Food Frequency Questionnaires (FFQ) were provided by the CFRI team to determine maternal dietary intakes during gestation. Administration of the F F Q took place at 16 and 36 weeks of gestation. The F F Q included sixty-seven food items grouped into nine groups: Dairy products; Table/cooking fat; Snack items; Breads, cereals and baked goods; Combination foods/meals; Meats and poultry; Fish; Vegetables and Fruits; and Beverages. During the administration of the F F Q , participants were asked to report their consumption of foods during the previous month. Intake of food items was evaluated in a daily, weekly, or monthly basis and usual portion sizes were recorded. Three-dimensional food models were also use to aid the participant in describing portion sizes. A.2 Maternal Supplement Intakes Participants were asked to disclose the use of any supplements taken during their gestational period. The brand name of the supplement, dosage, the date in which supplementation started, and the reason of taking them were also provided by the CFRI team. A.3 Data Analysis A.3.1 Dietary Intakes using the ESHA Food Processor Maternal dietary information obtained from the F F Q was entered into a nutrient database E S H A Food Processor version 8.3.0, 2004, using Canadian food products. Dietary information was entered on a weekly basis based on the number of servings and frequency of consumption. For example, i f a participant consumed 200g of tuna three times per month, her weekly intake would be calculated by multiplying the serving size (200g) by the frequency of consumption (3 times) and dividing this by the time period to transform the data into a weekly basis (30/7 weeks per month); therefore, the resulting weekly intake would be 140g of tuna. After completion, the weekly intake was transformed into daily intake values, by dividing the weekly value results by 7 days/week. A food list was saved for each F F Q at both 16 and 36 weeks of gestation, so all participants had two food lists. 92 Daily intakes were obtained for macronutrients and vitamins A , E and C by selecting the option of Mul t i -Column analyses. Vitamins were analyzed separately including intakes of supplements. The nutritional value of the supplements was found according to the brand name and the information on respective labels. Special consideration was taken for vitamins A and E, since resulting daily intake values from the database needed to be converted according to Dietary Reference Intake (DRI) units. A.3.2 Calculating Vitamin A Intake According to the Canadian DRI values, Vitamin A intake is expressed in micrograms of Retinol Activi ty Equivalents (ug R A E ) , which contemplates the vitamin A activity of plant and animal sources, mixed foods, and supplements. In order to calculate daily intake of vitamin A , conversions were necessary for both food and supplement sources. Total vitamin A activity was calculated by adding R A E from food and supplements. Dai ly intakes of vitamin A resulted from the nutrient database were shown as micrograms of Retinol Equivalent (RE). R E from food sources were divided into plants as source of carotenoids, and animal and mixed foods as source of retinol. According to the E S H A database, 1 ug of R E from retinol is equal to 1 ug of R A E , while 1 ug of R E from carotenoids is equal to 0.5ug of R A E , as shown in Table A . l . The calculated R A E from foods was summed up. Supplemental vitamin A was comprised of P-carotene and vitamin A acetate. To calculate R A E ' s , micrograms of P-carotene were multiplied by a factor of 0.5 and micrograms of vitamin A acetate was multiplied by 0.87. Table A .2 shows the conversion factor for supplements. Table A . 1 Conversion factor for food sources Source (ug) Vitamin A (jig RE) Vitamin A (ng RAE) P-carotene 6.00 1.00 0.50 Retinol l.OO 1.00 1.00 Other carotenoids a (~i n n n A n 1 12.00 1.00 0.50 Source: E S H A Food Processor Manual. Appendix A-14: Conversions Vitamin A . 93 Table A .2 Conversion factor for supplemental vitamin A a Source (1 jmg) Vitamin A (ug RAE) International Units (IU) P-carotene O50 ToT Vitamin A (retinol) 1.00 3.33 Vitamin A acetate 0.87 2.89 Vitamin A palmitate 0.55 1.82 Source: Health Canada: Drug and Health Products Monographs. A3.3 Calculating Vitamin E Intake For vitamin E, as stated by the Canadian DRI values, intake is expressed in milligrams of a-Tocopherol Equivalents (mg cx-TE), which includes food and supplement sources. According to the E S H A nutrient database, vitamin E intakes from food sources are calculated and presented as a-tocopherol. Therefore, no further calculation was necessary to convert a-tocopherol into a -TE, since the conversion factor is 1:1 ( E S H A Food Processor Manual. Appendix A-15). On the other hand, the vitamin E form in pre-natal supplements was a-tocopheryl acetate. Vi tamin E as a-tocopherol was calculated by multiplying the milligrams of a-tocopheryl acetate by the factor of 0.45! For example, a supplement intake of 30mg/d a-tocopheryl acetate was equal to 13.50mg/d a-tocopherol. Total vitamin E intake was calculated by adding up milligrams of a-tocopherol intake from food and supplements. Since the daily intake value represents only a-tocopherol, a factor of 1:1 was used to convert into a-TE. A.3.4 Calculating the Acceptable Macronutrient Distribution Range (AMDR) Intake of macronutrients were converted into energy and expressed as percent of energy (e.g. from a maternal daily intake of 2000kcal/d, 30g/d of total fat were eaten; so it represents 270kcal/d, and thus 13.5% of energy). Energy values are for fat and fatty acids 9kcal/g and for protein and carbohydrates 4kcal/g. Results were compared to the Acceptable Macronutrient Distribution Range ( A M D R ) . 94 A .4 Statistical Analyses Mann-Whitney U tests were used to test for differences in maternal dietary intakes between 16 and 36 weeks of gestation. Data was evaluated with descriptive statistics such as mean, standard error of the mean (SEM) , median, and quartiles. Results of intakes were compared to the Canadian Dietary Reference Intakes (DRI). Analyses were done using Microsoft® Excel 2002 and SPSS© (version 10.0). The level of significance used was p<0.05. 95 APPENDIX B: HPLC METHOD FOR DETERMINATION OF VITAMINS A AND E IN H U M A N M I L K B.l Human Milk Samples Human milk samples were provided by volunteers who were lactating at the time of this preliminary experiment. M i l k expression was done either manually or with a pump after nursing the baby for a 3 to 5 minute period (hindmilk). Human milk samples were then frozen in a sterile container in a home freezer prior to be transported to a -80°C freezer in Food Science Laboratory. M i l k samples were kept frozen at -80°C until being analyzed. B.2 Chemicals and Equipment Standards for H P L C all-Jra«s-retinol, a-tocopherol, 5-tocopherol and y-tocopherol were purchased from Sigma-Aldrich. Stock solutions of standards were prepared in ethanol and kept at -20°C and used within one week. Solvents including hexane, ethanol and methanol were H P L C grade and were obtained from Fisher Scientific. Pyrogallol was obtained from Fisher Scientific and was dissolved in ethanol to 12%. Gaseous nitrogen had standard for medical use. Materials such as screw cap vials (# 5182-0714) and screw caps (# 5182-0717) were purchased from Agilent Technologies. Polyspring glass inserts (C4012-530 150uL) were obtained from National Scientific Co. The equipment used was the Thermolyne M a x i M i x II, Beckman CS-6 Centrifuge (GH-3.7 rotor), and the chromatographer Agilent 1100 H P L C (Agilent Technologies) with Chemstation software. The column used was Phenomenex SphereClone 5u ODS(2) (5um, 150 x 4.6mm). B.3 Milk Fat Extraction M i l k samples were pre-treated by fat extraction. Methodology used for this purpose was described by Rodas Mendoza et al. (2003). A volume of 500uL of human milk was placed in a glass tube and 400uL 12% pyrogallol was added and vortexed for few seconds. Mixture was dissolved in l m L distilled water and 600uL ethanol and re-vortexed for 2min. Fat extraction was done twice. A total o f 4mL hexane was added and mix was centrifuged (2000 x g for lOmin). The organic layer from the sample extract was separated and a total of 96 3mL was saved and evaporated under nitrogen at ambient temperature. The fatty residue was reconstituted in 200uL of methanol, mixed for 5min, and transferred into an inserter in a vial. Vitamins A and E contents in human milk (all-frvmy-retinol, a-tocopherol, 8-tocopherol, and y-tocopherol) were analyzed by H P L C (See Section B.4 for method parameters). The injection volume was lOOuL. M i l k samples were analyzed in duplicate. B.4 Determinat ion of H P L C Method Parameters Vitamins A and E present in milk samples were initially identified by retention times and chromatographic peaks of external standards (Figure B . l and B.2). Detection was performed by a diode-array detector (DAD) with the full range of absorbance wavelength in order to determine the best wavelength for each vitamin isomer. The best absorbance wavelength was determined for each isomer after peak integration of the standards using the "select peak apex spectrum" feature from the Chemstation software for H P L C . LWU1 A, S r j - K i . l b MB-4WI |Aia MA.rWVSKJBM2Cn.IJp irnAU 125 4 so I 25 ' (1) IJIAU 60-20 <2> A A A 3 fflAU | 80-| 60 i 40 4 20 j 0 3 PI m f\ A A CMJP1 D. Sg=2S8.1CRef=of) (ATS MAY16>J032-i201.Op Figure B . l Chromatogram corresponding to standards (mixed) Peak identification: (1) all-^ratts-retinol; (2) a-tocopherol; (3) 8-tocopherol; (4) y-tocopherol 97 DAD1 A, Sig=325 S (1) mAU DAD1 C. Sig=298 mAU Figure B.2 Chromatogram corresponding to a milk sample Peak identification: (1) all-frara-retinol; (2) a-tocopherol; (3) 5-tocopherol; (4) y-tocopherol Detection of a-tocopherol was set at 292nm; 8-tocopherol and y-tocopherol at 298nm; and alWrarcs-retinol at 325nm. The run time required was less than 5min, but the method was set at 8min to allow the column to be washed for the remaining time (Table B . l ) . Table B . l H P L C parameters for determination of vitamin A and E isomers Parameter Conditions Wavelength a-tocopherol 292nm Wavelength 5-tocopherol 298nm Wavelength y-tocopherol 298nm Wavelength all-frvms'-retinol 325nm Mobi le phase 100% methanol F low rate 1.3mL/min Temperature 30°C Run time 8min 98 B.5 Vitamin Concentration in Milk Samples T o calculate v i t a m i n concentration i n m i l k samples, the resul t ing peak area [ m A U * s ] needed to be w i t h i n the range o f the ca l ibra t ion curves o f external standards. Concentrat ions o f v i tamins A and E were calculated by us ing the regression equation o f the cal ibrat ion curves, y = mx + b, i n w h i c h y corresponds to the peak area and x to the concentration o f the v i t amin (ug). The concentrat ion o f the v i t amin (ug) was mu l t i p l i ed by the convers ion factor o f (2 * 4/3), s ince 2 0 0 u L was recovered f rom the dry fatty residue, w h i c h was obtained from 3 out o f 4 m L o f the hexane layer. The result ing value was d i v i d e d by the in i t i a l m i l k vo lume o f 5 0 0 u L . The f ina l concentrat ion was expressed as u g / m L m i l k . The v i t amin E activity i n human m i l k was calculated accord ing to the formula ( u g / m L a-tocopherol x 1.0) + ( u g / m L 8-tocopherol x 0.03) + ( u g / m L y-tocopherol x 0.1) (Institute o f M e d i c i n e , 2000), and expressed as u g / m L m i l k . B.6 Linearity of Calibration Curves of Vitamin Standards Cal ib ra t ion curves o f external standards were obtained for each isomer by compar ing the peak area [ m A U * s ] versus respective k n o w n concentrations. F i v e standard solutions were made for each o f the tocopherol isomers and s ix for ret inol . The ranges i n concentration o f standards were a- tocopherol 2 . 5 - 5 0 u g / m L , 5-tocopherol and y- tocopherol 0 . 5 - 2 5 u g / m L and a l l - tomy-re t inol 0 . 0 5 - 2 . 5 u g / m L . A vo lume o f 2 0 u L was injected onto the H P L C co lumn . Ca l ib ra t i on curves were plotted by compar ing peak area and concentration o f v i t amin standard, [ m A U * s ] vs [ug]. The l ineari ty o f standard curves was expressed i n terms o f the determination coefficient ( R ), obtained f rom the regression equations f rom plots o f the integrated peak area [ m A U * s ] versus concentrat ion o f the standard [ug] (Figures B . 3 to B . 6 ) . The determination coefficient ( R ) o f the standard curves o f a l l v i t a m i n isomers was calculated for the three independent runs. Regress ion analyses resulted i n satisfactory l ineari ty, R 2 >0.990, as presented i n Table B . 2 . 99 y = 381.4002x-1.3404 0 0.1 0.2 0.3 0.4 0.5 0.6 [ug] Figure B.5 Calibration curve for y-tocopherol 100 180 160 140 120 100 80 60 40 20 0 Figure B.6 Calibration curve for all-frvms-retinol Table B.2 Regression equations and R 2 of calibration curves for vitamins A and E Vitamin isomer Run Regression equation a R2 a-tocopherol 1 y = 239.08x+ 1.19 1.0000 2 y = 239.32x4-0.64 1.0000 3 y= 239.58x4-0.68 1.0000 Average 1.0000 5-tocopherol 1 y = 288.93x 4-0.76 1.0000 2 y = 286.89x 4- 0.67 1.0000 3 y = 289.29x 4-0.57 0.9999 Average 0.9999 y-tocopherol 1 y = 381.40x- 1.34 0.9997 2 y = 378.18x- 1.21 0.9999 3 y = 378.83x-0.55 1.0000 Average 0.9998 alWrara-retinol 1 y = 3294.89x - 0.90 0.9999 2 y = 3391.52x- 1.22 0.9997 3 y = 3388.81x- 1.20 0.9998 Average 0.9998 From peak area [mAU*s] and concentration [ug] 101 y = 3294.8931x - 0.8995 F^0T9999" 1 0.01 0.02 0.03 0.04 0.05 0.06 [ug] B.7 Recovery To verify the reliability of the fat extraction methodology, two different human milk samples were spiked with an internal standard composed o f a mixture of vitamins A and E isomers. Four mixtures were used (e.g. M l , M 2 , M 3 and M4) . The composition is presented in Table B.3. Vi tamin A and E isomers had an initial concentration of lOOug/mL. Human milk (500uL) was spiked with 200uL of M l , M 2 , M 3 or M 4 , one mixture at a time. Table B.3 Composition of mixtures used for recovery analyses a-tocopherol 8-tocopherol Y-tocopherol all-fra/ts-retinol (uL) (_L) (uL) (uL) Ml 600 300 150 75 M2 75 600 300 150 M3 150 75 600 300 M4 300 150 75 600 Spiked samples were extracted in triplicate as described in Section B.3. Sample A was spiked with mixtures M l and M 2 and sample B was spiked with mixtures M 3 and M4 . Recovery was calculated by subtracting the peak area [mAU*s] of the non-spiked milk from the spiked milk. The resulted experimental peak areas corresponded to that of the added isomer standards in the mixture. Concentrations were calculated for each vitamin isomer in micrograms [pg] by using new calibration curves with a range of concentrations of those isomer standards used to make the mixtures. Experimental concentration (CE) was compared to the theoretical concentration (CT) of each isomer standard in the mixture added to the milk sample, and recovery was reported in percent (%). Table B.4 presents the percentages of recovery obtained after milk samples were spiked with internal standards, extracted, and analyzed by the H P L C . Recoveries were 99.72 ± 1.90% for a-tocopherol, 100.53 ± 2.32% for 8-tocopherol, 100.06 ± 2.32% for y-tocopherol, and 99.56 ± 1.26% for aM-trans-retinol. 102 Table B.4 Recovery of vitamin isomers added to milk samples A and B Milk Theoretical and Experimental Concentration [u.g] 0 Sample a-Toc 5-Toc y-Toc retinol A + Ml O r 5.33 2.67 1.33 0.67 C E 5.27 ± 0.04 2.62 ±0.01 1.33 ±0.01 0.66 ± 0.00 Recovery% b 98.79 ±0.70 98.41 ±0.55 99.81 ± 1.05 98.38 ±0.38 A + M2 C T 0.67 5.33 2.67 1.33 C E 0.68 ±0.01 5.53 ±0.02 2.67 ±0.01 1.35 ±0.01 Recovery% 101.66 ±1.19 103.62 ±0.43 100.20 ±0.22 101.04 ±0.44 B + M3 C T 1.33 0.67 5.33 2.67 C E 1.32 ±0.04 0.67 ±0.01 5.38 ±0.10 2.67 ±0.03 Recovery% 99.10 ±2.73 100.17 ± 1.48 100.97 ± 1.90 100.07 ± 1.17 B + M4 C T 2.67 1.33 0.67 5.33 C E 2.65 ± 0.04 1.33 ±0.03 0.66 ±0.00 5.25 ±0.03 Recovery% 99.32 ± 1.66 99.90 ± 2.25 99.25 ±0.59 98.40 ±0.59 Average Recovery % b 99.72 ± 1.90 100.53 ±2.32 100.06 ±2.32 99.56 ± 1.26 3 Concentration results in [ug] are shown as mean ± SD b Recovery % results are shown as mean ± SD d = Theoretical concentration of added vitamin standard C E = Experimental concentration of added vitamin standard after fat extraction and H P L C analysis 103 APPENDIX C: DATA ANALYSIS, FORMULAS, AND STATISTICAL RESULTS C l Box Plots C . l . l Interpretation of Box Plots B o x plots show the median value as the hor izonta l bar inside the box , the first quartile or 2 5 t h percentile ( Q l ) , the third quartile or 7 5 t h percentile (Q3) , and the whiskers extending from both ends o f the box w h i c h represent the values 1.5 box-lengths f rom the 2 5 t h or 75 l percentile, respect ively, or ± 2 . 6 9 8 times the standard deviat ion from the median (Ker r et a l . , 2002). The representation o f the box plot compared to the probabi l i ty density function o f the normal dis t r ibut ion curve is shown i n Figure C l . The box represents the interquartile range ( I Q R ) w h i c h is the range between Q l and Q 3 , and represents 5 0 % o f the sample. F igure C l C o m p a r i s o n o f box plot w i th normal dis t r ibut ion curve 104 C.1.2 Comparison of Milk D H A Contents Some w o m e n not taking D H A supplements dur ing gestation had m i l k D H A contents, at month 1 postnatally, s imi la r to those o f participants tak ing the supplement during gestation, 0.08-1.11 vs 0 .10-0 .72 g/ lOOg F A , respectively. In addi t ion, it can be seen that m i l k D H A contents f rom 2 participants w h o ment ioned not taking any D H A supplement during gestation were higher than the other group o f w o m e n (Figure C.2) . W i t h this informat ion the fact that w o m e n were either taking D H A supplements or were having h igh fish intakes remained inconc lus ive . 1.2 1.0 o < Q .6 .2 0.0 30 N D * o mm ™ 30 Yes Women taking DH^ supplement Figure C . 2 B o x plot compar ing m i l k D H A contents at mon th 1 postnatally ( « = 6 0 ) o Denotes values greater than ± 2 . 6 9 8 times the standard devia t ion from the median * Denotes values greater than ± 3 . 0 times the standard devia t ion f rom the median C.2 Antioxidant Capacity, Vitamin, and Fatty Acid Characteristics of Human Milk C.2.1 Multiple Regressions Model 1: Independent Var i ab le s : A lpha- tocophero l , Re t ino l , D H A , A A , E P A , L A , A L A Dependent V a r i a b l e : T A C ( O R A C ) 105 Table C l Summary of model 1, showing R 2 Model Summary6 Adjusted Std. Error of Model R R Square R Square the Estimate 1 .478a .229 .125 .5003 a- Predictors: (Constant), ALA, EPA, RETINOL, AA, ALPHA, LA, DHA b- Dependent Variable: ORAC Table C.2 Significance of the model ANOVAb Model Sum of Squares df Mean Square F Sig. 1 Regression 3.864 7 .552 2.205 .049a Residual 13.017 52 .250 Total 16.882 59 a- Predictors: (Constant), ALA, EPA, RETINOL, AA, ALPHA, LA, DHA b- Dependent Variable: ORAC Table C.3 Beta coefficients and significance of independent variables Coefficients3 Model Unstandardized Coefficients Standardi zed Coefficien ts t Sig. B Std. Error Beta 1 (Constant) 2.763 .458 6.038 .000 ALPHA .327 .093 .541 3.533 .001 RETINOL -2.068 2.023 -.153 -1.022 .311 DHA -.308 .496 -.103 -.622 .536 AA .513 .825 .080 .622 .537 EPA -.578 1.394 -.066 -.415 .680 LA -6.73E-03 .027 -.040 -.248 .805 ALA 5.120E-02 .158 .054 .323 .748 a. Dependent Variable: ORAC 106 Model 2 Independent Variables: Alpha-tocopherol, Retinol, co-3 F A (n3), co-6 F A (n6) Dependent Variable: T A C ( O R A C ) Table C.4 Summary of model 2, showing R Model Summary 6 Adjusted Std. Error of Model R R Square R Square the Estimate 1 .456a .208 .151 .4930 a. Predictors: (Constant), RETINOL, N6, ALPHA, N3 b. Dependent Variable: ORAC Table C.5 Significance of model 2 ANOVA" Model Sum of Squares df Mean Square F Sig. 1 Regression 3.513 4 .878 3.613 .011a Residual 13.369 55 .243 Total 16.882 59 a. Predictors: (Constant), RETINOL, N6, ALPHA, N3 b- Dependent Variable: ORAC Table C.6 Beta coefficients and significance of independent variables Coefficients3 Model Unstandardized Coefficients Standardi zed Coefficien ts t Sig. B Std. Error Beta 1 (Constant) 2.882 .379 7.607 .000 N3 -5.32E-02 .115 -.069 -.463 .645 N6 5.546E-03 .024 .034 .229 .820 ALPHA .307 .089 .507 3.436 .001 RETINOL -1.850 1.973 -.136 -.937 .353 a. Dependent Variable: ORAC 107 C.2.2 Bivariate Correlations Table C.7 Bivariate relationships, Pearson correlation coefficients, and significance ORAC ALPHA RETINOL DHA AA EPA LA ALA PUFA LCN3 LCN6 N3 N6 Ul ORAC Pearson Correlation 1.000 .439" .155 -.123 .003 -.057 -.097 -.110 -.108 -.101 -.017 -.137 -.093 -.054 Sig. (2-tailed) .000 .238 .349 .979 .668 .461 .404 .409 .444 .897 .298 .481 .682 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 ALPHA Pearson Correlation .439" 1.000 .557" -.021 -.076 .094 -.180 -.253 -.199 -.005 -.093 -.209 -.181 -.190 Sig. (2-tailed) .000 .000 .873 .562 .476 .168 .051 .127 .970 .477 .109 .166 .146 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 RETINOL Pearson Correlation .155 .557" 1.000 -.054 .038 -.048 -.043 -.145 -.054 -.060 .120 -.138 -.033 .000 Sig. (2-tailed) .238 .000 .684 .774 .718 .747 .269 .681 .646 .362 .292 .802 .997 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 DHA Pearson Correlation -.123 -.021 -.054 1.000 .187 .621" .183 .316* .285* .974" .229 .634" .189 .295* Sig. (2-tailed) .349 .873 .684 .153 .000 .162 .014 .027 .000 .079 .000 .148 .022 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 AA Pearson Correlation .003 -.076 .038 .187 1.000 .150 .240 .180 .299* .249 .774" .243 .286* .305* Sig. (2-tailed) .979 .562 .774 .153 .252 .065 .168 .020 .055 .000 .061 .027 .018 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 EPA Pearson Correlation -.057 .094 -.048 .621" .150 1.000 .005 .078 .078 .758" .109 .362- .011 .099 Sig. (2-tailed) .668 .476 .718 .000 .252 .971 .555 .555 .000 .409 .005 .931 .451 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 LA Pearson Correlation -.097 -.180 -.043 .183 .240 .005 1.000 .632** .987" .156 .379" .575" .998" .859* Sig. (2-tailed) .461 .168 .747 .162 .065 .971 .000 .000 .234 .003 .000 .000 .000 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 ALA Pearson Correlation -.110 -.253 -.145 .316* .180 .078 .632" 1.000 .721" .293* .171 .929- .621" .697* Sig. (2-tailed) .404 .051 .269 .014 .168 .555 .000 .000 .023 .190 .000 .000 .000 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 PUFA Pearson Correlation -.108 -.199 -.054 .285* .299* .078 .987" .721" 1.000 .263* .432- .689- .988" .884* Sig. (2-tailed) .409 .127 .681 .027 .020 .555 .000 .000 .042 .001 .000 .000 .000 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 LCN3 Pearson Correlation -.101 -.005 -.060 .974*' .249 .758" .156 .293* .263* 1.000 .272* .626- .166 .267* Sig. (2-tailed) .444 .970 .646 .000 .055 .000 .234 .023 .042 .036 .000 .204 .039 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 LCN6 Pearson Correlation -.017 -.093 .120 .229 .774" .109 .379" .171 .432" .272* 1.000 .244 .437" .387* Sig. (2-tailed) .897 .477 .362 .079 .000 .409 .003 .190 .001 .036 .061 .000 .002 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 N3 Pearson Correlation -.137 -.209 -.138 .634" .243 .362" .575" .929" .689- .626" .244 1.000 .570" .672* Sig. (2-tailed) .298 .109 .292 .000 .061 .005 .000 .000 .000 .000 .061 .000 .000 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 N6 Pearson Correlation -.093 -.181 -.033 .189 .286* .011 .998" .621" .988" .166 .437" .570" 1.000 .860* Sig. (2-tailed) .481 .166 .802 .148 .027 .931 .000 .000 .000 .204 .000 .000 .000 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 Ul Pearson Correlation -.054 -.190 .000 .295* .305* .099 .859" .697" .884" .267* .387- .672" .860" 1.000 Sig. (2-tailed) .682 .146 .997 .022 .018 .451 .000 .000 .000 .039 .002 .000 .000 N 60 60 60 60 60 60 60 60 60 60 60 60 60 60 **• Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed). C.2.3 Plots of Bivar iate Correlat ions 6 °1 . . 1 0.0 .1 .2 .3 Mlk alMrans-retinol (ug/rrL) Figure C.3 Relationship between milk a-tocopherol and all-^ra«5-retinol contents, r=0.557 (p<0.001) (w=60) 1.2 0.0] 0.0 .1 .2 .3 .4 Mlk EPA (g/100g FA) Figure C.4 Relationship between milk D H A and E P A contents, r=0.621 (p<0.001) («= 1.2 0.0| 0.0 .5 1.0 1.5 2.0 2.5 3.0 3.5 MlkALA(cyi00g FA) gure C.5 Relationship between milk D H A and A L A contents, r=0.316 (p<0.05) (n 24 22 0.0 .5 1.0 1.5 2.0 2.5 3.0 3.5 MlkALA(g/100gFA) Figure C.6 Relationship between milk L A and A L A r=0.632 (pO.OOl) («=60) C.3 Oxidative Stability of Human Milk C.3.1 Calculating Tertiles Tertiles showing low (I), medium (II), or high (III) levels were calculated using z-scores to standardize the distribution (Portney and Watkins, 2000) of milk variables. As shown in Figure C.6, both z-scores Z\ and Z2 were calculated to find an area lower than 1/3 and higher than 2/3 of that of the distribution curve. Z - scores were z\ = -0.4308 and Z2= 0.4308. z i z 2 Figure C.l Standardized distribution curve, showing tertiles areas and z-scores Tertiles were calculated for milk T A C values and contents of vitamins A and E, individual fatty acids, and unsaturation index (UI) of milk, by using the mean and standard deviation of the studied milk variables (Formula (c.l)), in which z is z i = -0.4308 or Z2= 0.4308, Y is the value in which the tertile ends, y is the mean, and s is the standard deviation. z = 7__y_ (c.l) s C.3.2 Calculating Odds Ratio (OR) To calculate O R , the odds of an event was defined as the ratio of the number of participants experiencing the event to the number of participants not experiencing the event while having been exposed to a factor. The event in this particular study is presenting a low 111 or high T A C in human milk, and the factors were low, medium, and high levels of vitamins, individual fatty acids, and unsaturation index of milk. The tabular formulation to determine O R is shown in Table C.8. The O R is the ratio of the odds of those participants presenting the event having been exposed to the factor (a / c) to the odds of participants not presenting the event while exposed to the factor (b / d). Table C.8 Tabular formulation of odds ratio (OR) Experiencing the Event Yes N o Exposed to Yes a b Factor No c d The formula to calculate O R is: O R = ale = a x d bid bxc (c.2) However, since sample size of this study is smaller than 100 («=60), it was necessary to add 0.5 to each number in the formula (Garb, 1996). The O R formula used is: O R = (a + 0.5) x (d+0.5) (6 + 0.5) x (c + 0.5) (c.3) A 2-way contingency table analysis was used to calculate O R from an online statistical database <http://statpages.org/ctab2x2.html>. C.3.3 Interpret ing Odds Rat io (OR) For the interpretation of a statistical significance in O R , it is also important to consider the confidence intervals (CI) and p values. The O R can take any positive value 112 above zero. The O R above 0 and below 1.0, can be interpreted as a reducing effect; an O R equal to 1.0 is consistent with no real effect; an O R above 1.0 is interpreted as increasing effect. The true O R value is given by the confidence intervals (CI). A t a 95% confidence, i f the CI does not include 1.0, then the result is significant at p value < 0.05 (Whitley and Bal l , 2002). For example, when evaluating the effect of a medium vitamin E content on causing a high T A C value in human milk, the O R is 2.23, the 95% CI is 0.74-6.80, and p=0.24; then the result is interpreted as the odds for having a high milk T A C value while having a medium vitamin E content is 2.23 times greater than i f milk has not a medium vitamin E content. The 95% CI shows that the odds can be both decreasing by 26% or increasing by 6.8 times the odds for having high milk T A C value, and it includes the O R value of 1.0, which suggests no real effect. The p value (p=0.24) confirms this interpretation, meaning that these findings can be occurring by 24% chance and thus are not statistically significant at p<0.05. C.3.4 Calcu la t ing Unsaturat ion Index (UI) The degree of fatty acid unsaturation, unsaturation index (UI), of human milk samples was calculated according to Formula (c.4), where mj is the number of double bonds and rj is the relative content of each unsaturated fatty acid in human milk samples (Gutierrez et al., 2005). UI (%) = Z (mj x rO (c.4) 113 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0101135/manifest

Comment

Related Items