Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Fractionation of expressed milk for the selective collection of hindmilk by mothers who deliver premature… Lalari, Vikki Valjeet 2002

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

Item Metadata

Download

Media
831-ubc_2003-0050.pdf [ 9.15MB ]
Metadata
JSON: 831-1.0090813.json
JSON-LD: 831-1.0090813-ld.json
RDF/XML (Pretty): 831-1.0090813-rdf.xml
RDF/JSON: 831-1.0090813-rdf.json
Turtle: 831-1.0090813-turtle.txt
N-Triples: 831-1.0090813-rdf-ntriples.txt
Original Record: 831-1.0090813-source.json
Full Text
831-1.0090813-fulltext.txt
Citation
831-1.0090813.ris

Full Text

FRACTIONATION OF EXPRESSED MILK FOR THE SELECTIVE COLLECTION OF HINDMILK BY MOTHERS WHO DELIVER PREMATURE INFANTS By VIKKI VALJEET LALARI B.Sc. Dietetics University of British Columbia, 1987 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE In THE FACULTY OF GRADUATE STUDIES (Human Nutrition) We pt this thesis as conforming the required standard UNIVERSITY OF BRITISH COLUMBIA December 2002 © Vikki Valjeet Lalari, 2002 UBC Special Collections - Thesis Authorisation Form Page 1 of 1 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the Un i v e r s i t y of B r i t i s h Columbia, I agree that the Li b r a r y s h a l l make i t f r e e l y a v a i l a b l e for reference and study. I further agree that permission for extensive copying of th i s thesis for s c h o l a r l y purposes may be granted by the head of my department or by his or her representatives. I t i s understood that copying or p u b l i c a t i o n of t h i s t hesis f o r f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of [ - J U r Y y i ^ r V u 4 ^ ( f ? V v A The U n i v e r s i t y of B r i t i s h Columbia Vancouver, Canada Date http://www.library.ubc.ca/spcoll/thesauth.html 2/10/00 Abstract Abstract Adequate weight gain is often difficult to achieve in premature infants due to illness, high energy requirements, and suboptimal intakes. Feeding strategies to enhance the energy and nutrient intake of premature infants fed small volumes are necessary for individualizing nutritional care and promoting optimal growth. One feeding strategy for premature infants fed maternal milk is to fractionate mothers' milk into foremilk (low fat milk) and hindmilk (high fat milk) with the subsequent feeding of the energy dense hindmilk fraction to promote weight gain. Published research on fractionating expressed milk for the selective collection of hindmilk is limited. The purpose of this study was to develop a fractionating protocol for the selective collection of hindmilk produced by mothers of premature infants. The basis of the fractionating protocol was to determine an acceptable method for separating expressed milk into foremilk and hindmilk. Three fractionating methods, based on the visual appearance of the milk, the milk expression time, and the volume of milk expressed, were compared for efficacy (change in fat content and adequacy of hindmilk volume) and practicality (subjective evaluation by mothers). The "Time" method was found to be the most acceptable and practical fractionating method for the selective collection of hindmilk. A regression equation for predicting the fat content of milk from the creamatocrit value (determined by the Creamatocrit Method) was developed using a large number of milk samples (n=155) with varying concentrations of fat. This equation, fat (g/100mL) = [0.572 x (creamatocrit value)] - 0.18, was found to be more suitable for estimating the fat content of milk in the present data set when compared to other published equations. The Crematocrit Method can be a useful clinical tool as part of a fractionating protocol for estimating the fat content of milk. This fractionating protocol is suitable for mothers who produce a sufficient quantity of milk (>125%) relative to their infant's prescribed enteral intake. The fractionating protocol developed in this study is applicable for a clinical or research setting as the basis of a hindmilk feeding strategy to promote weight gain in premature infants. ii Table of Contents TABLE OF CONTENTS A B S T R A C T ii TABLE O F CONTENTS iii LIST O F ABBREVIATIONS vii LIST O F TABLES viii LIST O F FIGURES x A C K N O W L E D G E M E N T S xi C H A P T E R 1. INTRODUCTION 1 1.1 Introduction 1 1.2 Study Overview 5 1.3 Purpose of the Study 5 1.4 Research Hypotheses 5 1.5 Primary Objectives 6 1.6 Specific Aims 6 1.7 Secondary Objectives 6 CHAPTER 2. REVIEW OF LITERATURE 8 2.1 Definition and Classification of the Premature Infant 8 2.2 Growth of the Premature Infant 9 2.3 Nutritional Risk Factors of Prematurity 12 2.4 Nutritional Management of Premature Infants 13 2.5 Enteral Nutrition and Premature Infants 15 2.6 Milk Production by Mothers Who Deliver Premature Infants 16 2.7 Fat Content of Human Milk 17 2.7.1 Inter-Individual Variation in the Fat Content of Milk 18 2.7.2 Gestational Age and Variation in the Fat Content of Milk 18 2.7.3 Stage of Lactation and Variation in the Fat Content of Milk 21 2.7.4 Throughout the Day (Diurnal) Variation in the Fat Content of Milk 22 2.7.5 Within a Feed Variation in the Fat Content of Milk 23 2.8 Hindmilk Feeding 25 2.9 Fractionating Human Milk into Foremilk and Hindmilk 27 2.10 Creamatocrit Measurements and Clinical Practice 28 2.11 Limitations to Our Present Knowledge 29 CHAPTER 3. DESIGN AND METHODS 31 3.1 Study Design 31 3.2 Development of Nutrition Study Booklet 32 3.2.1 Milk Expression Diary 32 3.2.2 Expressed Breast Milk Practices Survey 32 3.2.3 Evaluation 33 3.2.4 Socio-Demographic Questionnaire 33 iii Table of Contents 3.3 Study Participants 34 3.3.1 Participant Recruitment 34 3.3.2 Participant Selection Recruitment 34 3.4 Ethics 34 3.5 Study Procedures 35 3.5.1 Phase I: Development of a Protocol for Fractionating Expressed Human Milk into Foremilk and Hindmilk 35 3.5.2 Phase II: Reproducibility of a Protocol for Fractionating Expressed Human Milk into Foremilk and Hindmilk 36 3.5.3 Phase III: Determination of Expressed Breast Milk Practices 37 3.6 Laboratory Methods 37 3.6.1 Chemicals 37 3.6.2 Equipment 37 3.6.3 Lipid Analysis 38 3.6.3.1 Gravimetric Analyses 38 3.6.3.2 Creamatocrit 38 3.6.3.3 Direct Methylation/ Gas Liquid Chromatography (Fatty Acid Analysis) 39 3.6.4 Intra- and Inter- Assay Variability Measurements for Lipid Analysis 40 3.7 Data Analysis 41 CHAPTER 4. RESULTS 43 4.1 Participant Recruitment and Characteristics 43 4.1.1 Phase I - Development of a Protocol for Fractionating Expressed Human Milk into Foremilk and Hindmilk 43 4.1.2 Phase II - Reproducibility of a Protocol for Fractionating Expressed Human Milk into Foremilk and Hindmilk 47 4.1.3 Phase III - Determination of Expressed Breast Milk Practices and Volumes 51 4.2 Practices of Breast Milk Expression 55 4.2.1 Practices of Breast Milk Expression During the Baseline Period in Phase 1 55 4.2.2 Practices of Breast Milk Expression During the Baseline Period in Phase II 56 4.2.3 Practices of Breast Milk Expression in Phase III 57 4.3 Phase I - Development of a Protocol for Fractionating Expressed Human Milk into Foremilk and Hindmilk 58 4.3.1 Fat Content of Preterm Milk 58 4.3.2 Fat Content of Foremilk and Hindmilk 59 4.3.3 Volume of Expressed Preterm Milk 61 4.3.3.1 Volume of Expressed Preterm Milk During the Baseline Period 61 4.3.3.2 Foremilk and Hindmilk Volumes 61 4.3.4 Evaluation Summary of Fractionating Methods 64 4.3.4.1 Practicality Scores of the Fractionating Methods 64 iv Table of Contents 4.3.4.2 Preference Rankings of the Fractionating Methods 64 4.3.4.3 Mothers' Willingness to Perform the Fractionating Methods 67 4.3.4.4 Mothers' Comments on the Fractionating Methods 67 4.3.4.5 Summary for the Determination of an Acceptable Fractionating Method by Mothers 68 4.4 Phase II - Reproducibility of a Protocol for Fractionating Expressed Human Milk into Foremilk and Hindmilk 69 4.4.1 Fat Content of Preterm Milk 69 4.4.2 Fat Content of Foremilk and Hindmilk 70 4.4.3 Volume of Expressed Preterm Milk 71 4.4.3.1 Volume of Expressed Preterm Milk During the Baseline Period 71 4.4.3.2 Foremilk and Hindmilk Volumes 71 4.5 Phase III - Determination of Volume of Breast Milk Expressed by Mothers Who Deliver Premature Infants 74 4.5.1 Volume of Expressed Preterm Milk 74 4.5.2 Expressed Breast Milk Practices Survey 74 4.6 Relations Between the Modified Folch Method and the Creamatocrit Method 76 4.6.1 Correlation Between the Creamatocrit Value and the Total Fat Content of Expressed BreastMilk 76 4.6.2 Correlation Between the Creamatocrit Value and the Total Fat Content of Foremilk and Hindmilk Samples 77 4.6.3 Development of a Regression Equation for Estimating Total Fat Content of Expressed Breast Milk Using the Creamatocrit Method 78 4.7 Relations Between Fat Content of Preterm Milk and Milk Volume 80 4.8 Summary of Results with Respect to Hypotheses 81 CHAPTER 5. DISCUSSION 82 5.1 Introduction 82 5.2 The Fat Content of Preterm Milk 82 5.2.1 Inter-Individual Variation of the Fat Content of Milk 83 5.2.2 Determination of the Fat Content of Milk 85 5.3 Determination of an Acceptable Fractionating Method 86 5.3.1 Fat Content of Foremilk and Hindmilk 87 5.3.2 Volume of Foremilk and Hindmilk 89 5.3.3 Practicality Evaluation of the Fractionating Methods 91 5.3.4 Strengths and Limitations of the Fractionating Methods... 92 5.3.5 Summary of Results: Determination of an Acceptable Fractionating Method 94 5.3.6 Reproducibility of the Time Method in Phase II 95 5.4 Clinical Application of the Fractionating Protocol 96 5.4.1 Milk Production by Mothers Who Deliver Premature Infants: Considerations for Hindmilk Feeding 96 v Table of Contents 5.4.2 Creamatocrit Method as a Clinical Tool in the NICU 98 5.4.3 Fractionating Protocol for the Selective Collection of Hindmilk 100 5.4.4 Implications of Practice 100 5.5 Limitations of the Study 104 CHAPTER 6. CONCLUSIONS AND FUTURE DIRECTIONS 106 6.1 Conclusions 106 6.2 Future Directions 106 6.2.1 Future Directions for Research 106 6.2.2 Future Directions for Practice 108 R E F E R E N C E S 109 APPENDICES 119 Appendix 1: Pump Room Recruitment Letters 119 Appendix 2: Bedside Recruitment Letters 123 Appendix 3: Study Consent Forms 127 Appendix 4: Phase I Nutrition Study Booklet 134 Appendix 5: Phase II Nutrition Study Booklet 154 Appendix 6: Phase III Nutrition Study Booklet 171 Appendix 7: Conversion of Fatty Acid Content of Human Milk to Triglyceride and Total Fat Content 185 Appendix 8: Fresh EBM Creamatocrit Measurements: Relations Between Milk Fat Content (g/100mL) versus Creamatocrit Value 188 Appendix 9: Phase I and Phase II: Combined Data for Milk Fat Content and Volume of Expressed Milk 190 Appendix 10:Medium Chain Fatty Acid Content of Preterm Milk 192 Appendix 11 Phase I: Creamatocrit Measurements of Foremilk and Hindmilk 194 Appendix 12: Expressed Breast Milk Practices Survey Results 196 Appendix 13: Comparison of Selected Regression Equations for Estimating the Total Fat Content of Human Milk (g/100mL) from the Creamatocrit Value 204 vi List of Abbreviations LIST OF ABBREVIATIONS AAP American Academy of Pediatrics ANOVA analysis of variance AGA appropriate for gestational age BCCH British Columbia's Children's Hospital BPD bronchopulmonary dysplasia C carbon chain C P S Canadian Pediatric Society CV coefficient of variation d day EBM expressed breast milk ELBW extremely low birth weight E S P G N European Society of Pediatric Gastroenterology and Nutrition FT full term G L C gas liquid chromatography g gram h hour HMF human milk fortifier kcal calorie kg kilograms LGA large for gestational age L litre LBW low birth weight MCFA medium chain fatty acids mL milliliter min minute mo month NICU Neonatal Intensive Care Unit NPO nothing by mouth n number P statistical probability PT preterm S G A small for gestational age SCN Special Care Nursery SD standard deviation SEM standard error T G trigylceride uL microlitre VLBW very low birth weight wks weeks yr year vii List of Tables LIST OF TABLES Table 2.1 Recommended Daily Nutrient Intakes for Enterally Fed Stable Premature Infants 11 Table 2.2 Selected References on the Fat Content of Mature Milk Produced by Mothers of Preterm (PT) and Term (FT) Infants 19 Table 2.3 Selected References Summarizing the Within Feed Variation in the Fat Content of Human Milk 24 Table 3.1 Effect of Freezing Human Milk on the Creamatocrit Measurement 39 Table 3.2 Intra and Inter- Assay Variability Tests for the Modified Folch Method of Total Lipid Extraction and the Creamatocrit Method 40 Table 4.1 Characteristics of Participants in Phase 1 45 Table 4.2 Characteristics of Participants' Infants in Phase 1 46 Table 4.3 Characteristics of Participants in Phase II 49 Table 4.4 Characteristics of Participants' Infants in Phase II 50 Table 4.5 Participant Characteristics in Phase III 53 Table 4.6 Characteristics of Participants' Infants in Phase III 54 Table 4.7 Practices of Breast Milk Expression During the Baseline Period in Phase 1 55 Table 4.8 Practices of Breast Milk Expression During the Baseline Period in Phase II 56 Table 4.9 Practices of Breast Milk Expression in Phase III 57 Table 4.10 Fat Content of Preterm Milk in Phase 1 59 Table 4.11 Fat Content (g/1 OOmL) of Foremilk and Hindmilk Separated Using Three Fractionating Methods 60 Table 4.12 Volume of Expressed Preterm Milk During the Baseline Period 61 Table 4.13 Comparison of Volumes of Foremilk and Hindmilk Obtained Using Three Methods of Fractionating Expressed Milk 62 Table 4.14 Comparison of Practicality Score for Three Fractionating Methods 64 Table 4.15 Fat Content of Preterm Milk in Phase II 70 Table 4.16 Fat Content (g/1 OOmL) of Foremilk and Hindmilk Using the Time Method 70 Table 4.17 Volume of Expressed Preterm Milk During the Baseline Period 71 Table 4.18 Foremilk and Hindmilk Volumes Expressed Using the Time Method....72 Table 4.19 Volume of Expressed Preterm Milk in Phase III 74 viii List of Tables Table A1 Molecular Weight of Selected Fatty Acids and Gylcerol in Human Milk..186 Table A2 Molecular Weight (MWT) and T G Conversion Factor for Select Triglycerides in Human Milk 187 Table A3 Determination of the Total Fat Content of Preterm Milk 191 Table A4 Total Fatty Acid Content and MCFA content of Preterm Milk 193 Table A5 Phase I : Creamatocrit Measurements of Foremilk and Hindmilk 195 Table A6 Techniques Mothers Used to Stimulate Milk Production 197 Table A7 Mothers' Experiences with the Letdown Reflex 198 Table A8 Stress and Milk Production 201 Table A9 Comparison of Selected Regression Equations for Estimating the Total Fat Content of Human Milk (g/ 100mL) from the Creamatocrit Value 205 ix List of Figures LIST OF FIGURES Figure 4.1 Ability of Volume of Mothers' Expressed Hindmilk to Meet Infants' Intake 63 Figure 4.2 Frequency of Preference Ranking for Three Fractionating Methods.... 65 Figure 4.3 Comparison of Willingness to Perform Fractionating Method 66 Figure 4.4 Ability of Volume of Mothers' Expressed Hindmilk to Meet Infants' Intake (Phase II) 73 Figure 4.5 Ability of Volume of Mothers' Expressed Milk to Meet Infants' Enteral Intake 75 Figure 4.6 Relations Between Fat Content Values as Determined by the Modified Folch Method and the Creamatocrit Method 77 Figure 4.7 Relations Between Fat Content Values as Determined by the Modified Folch Method and the Creamatocrit Method in Foremilk and Hindmilk Samples 79 Figure 4.8 Scatterplot of Breast Milk Fat Content and Milk Volume Produced by Mothers of Premature Infants 80 Figure A.1 Relations Between Fat Content Values as Determined by the Modified Folch Method and the Creamatocrit Method 189 x ACKNOWLEDGEMENTS Acknowledgements I would like to express my sincere gratitude to the following people who have supported, encouraged and guided me over the last few years during my graduate studies. I am particularly thankful to my graduate supervisor, Dr. Sheila Innis, for providing me with a wonderful research opportunity involving premature infants and neonatal nutrition. Her nutrition knowledge, vision, and research expertise were much appreciated and provided me with insights into the research process. I would like to thank my thesis committee members, Dr. Ryna Levy Milne and Dr. John Symth for their support, guidance and kindness over the last few years. I am thankful for the support, guidance and time given by Dr. Susan Barr. I also want to thank Roger and Jannette for their patience, guidance and assistance with lab issues; in particular teaching me lipid analysis methodology. I am sincerely grateful to the Nutrition Services Department at B.C.'s Children's Hospital for granting me an education leave of absence and for supporting me as I pursued my Master's Degree. I also want to thank my fellow colleagues at B C C H , for their warm words of encouragement and support over the last few years. I would like to thank Carol, Jadine, Lucy and Cathy for being great friends and encouraging me throughout my graduate studies. To my fellow grad students from the Innis Lab, in particular, Carolanne, Loraina, Patty, Sylvia, Sandra, Tim, Angela, and Alice, thank-you from the bottom of my heart for your never ending support, guidance, encouragement and most of all friendship. I would like to thank Laurie Ainsworth who graciously offered statistical advice when I really needed it. I am indebted to the mothers who participated in my study and wish to thank them for their time, feedback, and generosity. Finally, I would like to thank my family for always being there for me in whatever challenges I take on. To Darci, Paul, Keira, Kailyn, Jaren, Mom and Dad, thank-you for your love, support and patience, especially during the last few years. xi Introduction CHAPTER 1 INTRODUCTION 1.1 INTRODUCTION The provision of optimal nutrition plays an important role in the survival, growth and development of premature infants. The American Academy of Pediatrics (AAP) recommends the optimal diet for premature infants as one that supports growth at intrauterine rates without imposing stress on the infant's immature metabolic and excretory functions (AAP 1977). The nutritional goal for premature infants is to provide adequate calories and essential nutrients for tissue accretion and growth similar to an intrauterine rate of 12-16 g/kg/day (Ehrenkranz 2000, Lubchenco 1963). Suboptimal weight gain is a common problem for premature infants. Reasons for poor weight gain include: inadequate energy and nutrient intake due to fluid restrictions, metabolic immaturities (e.g. limited absorption capabilities), high requirements (e.g. infants with chronic lung disease), nutrient losses due to feeding methods (e.g. fat adhering to feeding tubes), and prolonged periods of suboptimal nutritional support (e.g. due to fear of necrotizing enterocolitis). For enterally fed premature infants, appropriate growth and protein accretion have been estimated to require an energy intake of at least 120 kcal/kg/day (AAP 1985) and a protein intake of 3-4 g/kg/day (CPS 1995). Human milk is the recommended enteral feed for all infants, including premature infants (AAP 1997). The unique composition and characteristics of human milk offer many advantages to infants fed mother's milk. Feeding human milk provides immunological, nutritional and developmental benefits for the infant, as well as psychological benefits for the mother. The average fat content of human milk is 3.5 - 4.5 g/100 mL, representing about 40 - 50% of total energy (Sapsford 2000, Hamosh 1994). Fat is the most variable macronutrient in human milk. The fat content of milk varies within a feeding (from foremilk to hindmilk), during the day and with stage of lactation (Jensen 1989). A major source of variability, however, is between women (Hytten 1954e). The variability in the fat content of milk among women has been reported to range from less than 2 to greater than 5 g/100 mL (Anderson 1981,Hytten 1954a). This variation in the fat content of milk can result in variations in the energy density of the milk (assuming the protein and lactose content of the milk is relatively stable). A mother, who produces mature milk with a low fat content, is likely to also produce milk with a low energy density. Anderson et al. (1981) reported data on an individual mother whose 1 Introduction milk fat content (2.4-2.7g/100mL) was consistently low over a one-month study interval. The corresponding energy density of the milk was also low during the study interval, at approximately 50-60 kcal/100ml_. In comparison, the mean energy density of mature preterm milk in Anderson's study was approximately 70 kcal/100ml_; this value is similar to that reported by other investigators (Atkinson 1981, Butte 1984,Gross 1980). Coordination of the suck and swallow reflex does not develop until about 32-34 weeks gestation (Jones Wessel 2000). Because of this, premature infants who are enterally fed are dependent on tube feeds for nutrition support. In current practice, infants are fed a prescribed enteral volume and are assumed to obtain a set amount of calories based on the average fat content of human milk. The wide variation of human milk fat content suggests some infants fed by volume (where theoretically 50% of calories are from fat), receive less than adequate energy for optimal growth. Current feeding strategies to promote weight gain in infants receiving expressed human milk include: addition of human milk fortifier to increase the calorie, protein, vitamin and mineral density; addition of powdered term infant formula again to increase energy and essential nutrient density; addition of single macro-nutrients (e.g. polycose powder, promod powder, MCT oil); increasing the total milk intake; or supplementing the milk feed with a hypercaloric premature infant formula. Another feeding strategy is to fractionate expressed human milk into two fractions: foremilk and hindmilk, and selectively feed hindmilk to those infants with poor weight gain. The fat content of foremilk, milk at the beginning of a feed, has been reported to be 1.7-3 fold lower than hindmilk, milk later in a feed (Hall 1979, Neville 1984, Valentine 1994). Feeding hindmilk (higher fat content) may provide infants with higher energy intakes to promote weight gain and protein utilization. When fat intake, and thus calories are limited, protein may be used as an energy source resulting in negative nitrogen balance and inadequate lean body tissue growth (Micheli 1993, Noss 1999). The use of hindmilk to increase weight gain in premature infants supports and promotes the use of human milk, instead of using infant formula. There is limited published data on feeding hindmilk to promote weight gain in very low birth weight (birthweight <1500 g) infants. Valentine et al. (1994) reported improved weight gain in 15 low birth weight infants (birthweight <2500 g) fed fortified hindmilk (milk containing a higher fat concentration and energy density compared to composite milk). The rate of weight gain in the infants was measured before and after hindmilk feedings. The 2 Introduction investigators reported a significant increase (P<0.0001) in the rate of weight gain from WK-1 (usual milk feeding period) to WK-2 (hindmilk feeding period); a mean (± SD) increase of 7.0 ± 4.4 g/kg/day (Valentine 1994). The range of weight gain was -0.7-13 g/kg/day and 12-21 g/kg/day during WK-1 and WK-2, respectively. Several different methods have been described for separating expressed human milk for the selective collection of hindmilk. However, there is no standardized method for separating milk into foremilk and hindmilk. Valentine et al. (1994) separated expressed human milk based on a "time" method, where foremilk was defined as the milk fraction collected 2-3 minutes after milk flow began ("letdown") and hindmilk was defined as the milk fraction collected thereafter. Lang (1997) described a method of separating milk into foremilk and hindmilk based on a set "volume" or proportion of milk expressed at each pumping session. It is also possible to separate milk based on a change in the color and/or consistency of the milk. This method had not been reported in the literature; it relies on the mothers' observations for separating milk based on the visual appearance of the milk. Mothers of premature infants may express their breast milk for many weeks to months before exclusive breastfeeding is successfully established. Establishing and maintaining an adequate milk supply is a common challenge for mothers who deliver premature infants. The average milk volume expressed from mothers who deliver premature infants has been reported to be approximately 600 mL/day at 4 weeks postpartum, with a range of 43 - 1773 mL (Hopkinson 1988). Clearly, there is a wide variation among women in the volume of milk produced. Many factors may affect the volume of breast milk produced; these include, the frequency of breast pumping, the duration of breast pumping, maternal stress, medications, the type of breast pump (electric pump versus manual expression), the method of breast pumping (sequential versus simultaneous pumping of both breasts), maternal alcohol consumption and maternal cigarette smoking (de Carvalho 1985, Ehrenkranz 1986, Green 1982, Hopkinson 1988, Howard 1999, Hurst 1999, Jones 2001). Hopkinson et al. (1988) suggested that optimal milk production by mothers of premature infants was achieved when mothers expressed their milk at least 5 times per day and whose total pumping durations were greater than 100 minutes/day. The separation of mother's milk into foremilk and hindmilk can be considered for mothers who have a milk supply that is in excess of their infant's enteral intake. The 3 Introduction average milk volume produced by mothers of premature infants in the Special Care Nursery at British Columbia's Children's Hospital and the prevalence of factors that may influence their milk volume production are not known. The proportion of mothers of premature infants who are suitable candidates for fractionating their breast milk is also unknown. This study was designed to develop a protocol for fractionating human milk into foremilk and hindmilk that can be practically incorporated by mothers into their daily milk expression routine, that provides a sufficient quantity of hindmilk to meet their infant's prescribed enteral intake, and that results in at least a 1.5 fold increase in fat content from foremilk to hindmilk. This study was also designed to explore the potential variation of expressed breast milk volume produced by mothers of premature infants and document potential factors that may influence breast milk production in mothers who deliver premature infants. 4 Introduction 1.2 STUDY OVERVIEW This research was conducted in three sequential phases. Phase I - Development of a protocol for fractionating expressed human milk from mothers who deliver premature infants into foremilk and hindmilk. Phase II - Establishing the reproducibility of a protocol for fractionating expressed human milk from mothers who deliver premature infants into foremilk and hindmilk. Phase III- Examination of expressed breast milk volumes produced and practices of breast milk expression by mothers who deliver premature infants. Each phase of the study used a convenience sample of mothers whose premature infants were patients in the Special Care Nursery at B.C.'s Children's Hospital. The purpose of developing a fractionating protocol was to allow subsequent research to consider the nutritional benefits of feeding energy dense mother's milk to very low birth weight infants with poor weight gain. The examination of the practices of expressed breast milk expression and the milk volume expressed was conducted to support future studies that involve the expression of mother's milk for feeding premature infants. 1.3 PURPOSE OF THE STUDY To develop a practical protocol for the selective collection of hindmilk from expressed breast milk of mothers who deliver premature infants. 1.4 RESEARCH HYPOTHESES 1a) An acceptable method can be developed for separating the expressed milk of mothers who deliver premature infants into foremilk and hindmilk, which results in hindmilk with at least a 1.5 fold higher fat content than foremilk. 1b) An acceptable method can be developed for separating the expressed milk of mothers who deliver premature infants into foremilk and hindmilk, in which the volume of hindmilk collected will meet the infant's prescribed enteral intake. 1c) A method for fractionating and collecting hindmilk from the expressed milk of mothers who deliver premature infants can be developed which is acceptable and practical for mothers to incorporate into their daily milk expression routine. 5 Introduction 1.5 PRIMARY OBJECTIVES 1) To compare three methods for separating expressed breast milk into foremilk and hindmilk: the visual method, the time method and the volume method, and determine which of these is the most acceptable fractionating method. 2) To determine what proportion of mothers participating in this research would be suitable candidates for fractionating their breast milk into foremilk and hindmilk, based on the expression of a sufficient quantity of milk relative to their infant's prescribed enteral intake. 1.6 SPECIFIC AIMS 1a) To develop an acceptable protocol for fractionating expressed human milk to selectively collect hindmilk, based on the following criteria: (i) hindmilk containing at least a 1.5 fold higher fat content than foremilk; (ii) hindmilk volume meeting the prescribed enteral intake of the infant; and (iii) mothers' subjective evaluation of the practicality of the method, willingness to perform the method, and overall preference. 1 b) To demonstrate the preferred fractionating method is reproducible with a further, independent group of mothers, as determined by the ability of the expressed hindmilk volume to meet the infant's prescribed enteral intake and hindmilk containing at least a 1.5 fold higher fat content than foremilk. 2) To quantify breast milk expression volumes and to document breast milk expression practices of mothers participating in the study. 1.7 SECONDARY OBJECTIVES 1) To determine the range of fat content in breast milk expressed by mothers whose premature infants are patients in the Special Care Nursery at B.C.'s Children's Hospital. 2) To develop a predictive equation for estimating the fat content of expressed breast milk using the creamatocrit value, as determined by the Creamatocrit Method. 6 Introduction 3) To examine relations between the fat content of breast milk and the volume of breast milk expressed by mothers whose premature infants are patients in the Special Care Nursery at B.C.'s Children's Hospital. 4) To examine factors that may affect the breast milk production in mothers whose premature infants are patients in the Special Care Nursery at B.C.'s Children's Hospital. 7 Literature Review CHAPTER 2 LITERATURE REVIEW The nutritional management of premature infants is complex and challenging. It requires an understanding of the many nutritional risk factors associated with premature birth, the ability to apply neonatal nutrition guidelines and feeding strategies in clinical practice and the insight for individualizing nutritional therapy in order to meet the nutritional goals of the infant. Feeding strategies that promote the use of maternal milk and facilitate weight gain are essential in the nutritional management of premature infants. The following literature review focuses on key components involved in the nutritional management of premature infants and describes a feeding strategy that utilizes mother's milk to improve weight gain in premature infants. The first few sections of this chapter define and classify the premature infant, review growth of premature infants, and describe the nutritional risk factors associated with prematurity. The following sections of this chapter review the nutritional management of premature infants (in particular, enterally fed infants), the current feeding strategies available for premature infants, and the breast milk production of mothers who deliver prematurely. The next few sections of this chapter review the variations in the fat content of human milk, in particular the increase in fat content from foremilk to hindmilk that occurs during a feed or pumping session. The hindmilk feeding strategy (a strategy used to increase the infant's energy intake) is reviewed and potential fractionating methods for the collection of hindmilk (a high fat, energy dense milk fraction) are discussed. The last sections of this chapter review a simple and rapid method (the Creamatocrit Method) for estimating the fat content of milk in a clinical setting and discuss the limitations of our knowledge pertaining to methods for fractionating milk for the selective collection of hindmilk. 2.1 Definition and Classification of the Premature Infant A premature infant is defined as an infant born less than 37 weeks gestation. Infants are classified according to birth weight as low birth weight (LBW) infants, infants who weigh <2500 grams; very low birth weight (VLBW) infants, infants who weigh <1500 grams; and extremely low birth weight (ELBW) infants, infants who weigh <1000 grams. Infants can also be classified based on their size for gestational age at birth. 8 Literature Review Infants who are small for gestational age (SGA) have a birth weight less than the 10 th%ile. Infants who are appropriate for gestational age (AGA) have a birth weight between the 10th%ile and the 90 th%ile. Infants who are large for gestational age (LGA) have a birth weight greater than the 90 th%ile. Arbuckle et al. (1993) reported that 6.2% of all Canadian live births (1986-1988) were premature, 5.8% were low birth weight and 1.2% were very low birth weight (Arbuckle 1993). The incidence of premature births in British Columbia (1995) was 6.5% of live births and 5.3% for low birth weight infants (BC Vital Statistics 1997). Morbidity and mortality in premature infants are related to the infant's birth weight and gestational age. As birth weight and gestational age increase, neonatal mortality decreases (Anderson 2000). The survival of premature infants has improved over the last 20 years due to advances in medical management, technology and nutrition support. 2.2 Growth of the Premature Infant Growth is a key component in the nutritional assessment of premature infants. Comparison of an individual infant's growth parameters to established growth curves or expected growth rates are helpful in assessing growth status, monitoring growth trends and identifying growth abnormalities. Deviations from established growth curves or expected growth rates might indicate nutritional inadequacies or excesses. The ideal rate of growth for infants born prematurely is unknown. Intrauterine growth rates of 15g/kg/day have been recommended as the growth standard for premature infants (Ehrenkranz 2000, C P S 1995, Peirea 1995). The ability to achieve intrauterine growth rates may be an unrealistic expectation in some premature infants as they adjust to the stresses of extrauterine life, adapt to an exogenous nutrient supply and cope with various neonatal medical problems. Several researchers in the last few years have reported postnatal growth rates of premature infants (once birth weight was regained) to be similar to intrauterine rates (Ehrenkranz 1999, Pauls 1998, Wright 1993). The data used to generate postnatal charts generally represent a variety of medical and nutrition practices and may not reflect ideal growth. The growth status of premature infants is assessed and monitored by plotting anthropometric measurements (weight, length and head circumference) on intrauterine or extrauterine (postnatal) growth charts. Intrauterine growth charts are based on anthropometric data of infants born at different gestational ages (Anderson 2002). 9 Literature Review These cross-sectional growth charts represent estimates of intrauterine growth of the human fetus (Lubchenco 1963, Thureen 2000, Sparks 2000, Katrine 2000). There are several intrauterine growth charts available for clinical use in the neonatal intensive care unit (NICU). The most widely used intrauterine growth charts are the Luchenco and Babson/Benda growth charts (Katrine 2000). Intrauterine growth charts differ in the geographical location (in particular altitude) where the data were collected, the ethnicity of the population sampled, the sample size, and the year the data were collected (Anderson 2002, Sparks 1984). Postnatal growth charts reflect the longitudinal growth pattern of premature infants. Several postnatal growth charts have been developed in the last 55 years. These growth charts differ in the length of time infants were followed after birth (e.g. followed to 30 days of life or 120 days of life), the medical condition of the infants (e.g. medically uncomplicated or infants with major morbidities), the nutritional practices of the neonatal units (e.g. enterally fed only or receiving both parenteral and enteral nutrition), the sample size of the subjects recruited (e.g. 24 subjects to 1660 subjects), the location of the data collected (e.g. single site or location or multiple sites), the birth year of the infants (eg. born in the 1940s or mid 1990s) and the different ranges of birth weights (eg. 1000-2500g or 501 -1500 g) (Dancis 1948, Ehrenchranz 1999, Wright 1997, Brousius 1984, Pauls 1998). The postnatal growth chart by Dancis et al. (1948) is still widely used despite significant differences in neonatal care and nutrition practices since the chart was first published (Ehrenkranz. 2000). Ehrenkranz et al. (1999) have published the most recent postnatal growth charts, which represent data collected from 12 centers throughout the USA (Ehrenkranz 1999). Data were collected on 1660 infants born between 1994-1995 with birth weights ranging from 501 to 1500g. Growth measurements were collected until an age of 120 days, discharge, transfer, death or a body weight of 2000g. The growth charts reflect current aspects of neonatal care, in particular current nutritional practices (Anderson 2002, Ehrenkranz 1999). Ehrenkranz et al. (1999) found that after infants regained their birth weight (at approximately day 12-17), the average weight gain ranged from 13.5 to 16.3 g/kg/day. These results are similar to reported intrauterine weight gains of 15 g/kg/day or = 1.5%/day (Ehrenkranz 1999, Sparks 1984). Length and head circumference increments ranged from 0.9-1.0cm/wk and 0.7-1.0 cm/wk, respectively, and were similar to growth velocities reported by Luchenco et al. (1966) (Ehrenkranz 1999). 10 Literature Rev iew Nutritional strategies for premature infants are general ly a imed at achieving in utero rates of fetal growth. Energy intakes >120-130 kcal /kg/day and protein intakes of 3-4 g/kg/day (as well as other nutrients summar ized below) are recommended to ach ieve desired growth rates ( A A P 1998, C P S 1995, E S P G N 1987, Nevin-Fol in 2001). A summary of recommended daily nutrient intakes for stable enterally fed premature infants is presented in Table 2.1. Table 2.1 Recommended Daily Nutrient Intakes for Enterally Fed Stable Premature Infants 1 Nutrient Recommended Intake Energy (kcal/kg) 120-130 Protein (g/kg) 3-4 Fat (% total energy) 40-55% Carbohydrate (% total energy) 40-50% Calcium (mg/kg) 160—240 Phosphorus (mg/kg) 78-118 Magnesium (mg/kg) 4.8-9.6 Sodium (mEq/kg) 2.5-4.0 Potassium (mEq/kg) 2.5-3.5 Chloride (mEq/kg) 2.5-4.0 Iron (mg/kg) 3-4 (birthwt <1000g);2-3 (birthwt >1000g) Zinc (mg/kg) 1 Selenium (ug/kg) 1.3-3.0 Chromium (ug/kg) 0.1-0.5 Molybdenum (ug/kg) 0.3 Manganese (ug/kg) 0.75-7.5 Iodine (ug/kg) 30-60 Vitamin A (lU/kg) 700-1500 Vitamin D (lU/day) 400-800 Vitamin E (mg/kg) 0.5-0.9 Vitamin K (ug/kg) 8-10 Vitamin C (mg/kg) 6-10 Thiamin (mg/kg) 0.04-0.05 Riboflavin (mg/kg) 0.36-0.46 Niacin (NE/100kcal) 0.72 Pyridoxine (mg/g of protein) 0.015 Vitamin B 1 2 (ug/kg) 0.15 Folate (ug) 50 Pantothenic Acid (mg/kg) 0.8-1.3 Biotin (ug/kg) 1.5 1987 Adequate weight gain is often difficult to achieve in premature infants due to i l lness, high energy and nutrient requirements and subopt imal intakes. Car l son and Zeigler (1998) reported the nutrient intakes and growth of 51 very low birth weight (birthweight <1300g) infants during hospital ization. The authors reported nutrition and growth outcomes during four phases of nutritional management : the parenteral period (age 0-14 days), the transitional period (age 15-35 days), early enteral period (age 36-11 Literature Review 56 days), and the late enteral period (age 57 to term). Parenteral nutrition was provided until infants received an enteral intake of approximately 100ml_/kg/day. Enteral feedings were started between day 1 to 18 of life and consisted of either expressed breast milk (fortified with a commercial human milk fortifier once enteral volumes reached 80-100mL/kg) or a hypercaloric preterm infant formula. The mean energy and protein intake during the transitional, early enteral and late enteral periods ranged from 99-110 kcal/kg/day and 2.5 to 2.7 g pro/kg/day. The corresponding mean rate of weight gain during these three feeding periods ranged from 11.6-13.8 g/kg/day. The investigators reported that the slow weight gain observed in their study (relative to in utero growth rates) was primarily a consequence of suboptimal protein and energy intake. 2.3 Nutritional Risk Factors of Prematurity Premature infants are at risk for nutritional deficiencies and poor growth for many reasons. Nutritional risk factors include: poor nutrient stores at birth, increased nutrient requirements for rapid growth, immature organ function, medical conditions associated with prematurity and suboptimal nutrient intakes (Anderson 2002, Fulham 2000, Pereira 1995). Premature infants are born with low nutrient stores compared to term infants, as many nutrients such as fat, protein, vitamins and minerals, are accumulated during the last trimester of pregnancy (Anderson 1999, Wilson 1995). A premature infant weighing 1 kg (2.2 lb) has only 1-2 % of their body weight as fat, as compared to a term infant weighing 3.5 kg (7.7 lb) who has 15-16% body weight as fat (Heird 1972, Wilson 1995). A 1 kg premature infant is born with a non-protein caloric reserve (glycogen and fat stores) of approximately 110 kcal/kg (Heird 1972). These endogenous energy reserves become quickly depleted, unless an exogenous source of nutrients is provided. Early administration of nutritional support in the form of parenteral and /or enteral nutrition is essential to avoid caloric deprivation and protein catabolism. Theoretical projected survival times for a 1 kg premature infant has been estimated to be 4 days in a starved state (intravenous fluids containing no exogenous calories) and 11 days in a semi-starved state (intravenous fluids containing a 10% glucose solution) (Heird 1972, Atkinson 1994). Premature infants may have immature organ systems that limit the digestion, absorption and utilization of nutrients (Atkinson 1994). The functional and structural immaturity of the gastrointestinal (Gl) system has a major impact on how, what and 12 Literature Review when an infant is nourished. Low levels of digestive enzymes (i.e. lipase and lactase) and low concentrations of bile acids affect the absorption of nutrients from the Gl tract (Anderson 1999). Immature coordination of suck and swallow in infants less than 32-34 weeks gestation often necessitates the dependence on tube feedings for nutrition support until the infant is able to breastfeed or bottle feed efficiently (Jones Wessel 2000). Severe medical conditions, such as bronchopulmonary dysplasia (chronic lung disease), increase the infant's energy needs (Anderson 1999, Tsang 1993). Suboptimal parenteral and/or enteral intakes due to fluid restrictions, medical instability or feeding intolerance limit the amount of nutrients and calories supplied to the infant, thereby contributing to poor weight gain (Anderson 2002, Atkinson 1994). 2.4 Nutritional Management of Premature Infants Feeding strategies to optimize the calorie and nutrient intake of premature infants (especially VLBW) are essential to their nutritional management. Nutrition recommendations, nutrient requirements and feeding guidelines have been developed to assist clinicians in managing infants at high nutritional risk (AAP 1998, C P S 1995, E S P G N 1987, Groh-Wargo 2000,Tsang 1993). However, there is lack of agreement on what represents optimal nutrition for premature infants. Feeding practices and strategies used in the nutritional management of premature infants tend to vary from one neonatal unit to another (Evans 2000,Ziegler 2002). The American Academy of Pediatrics, AAP, recommends the "optimal diet for the low birth weight infant may be defined as one that supports a rate of growth approximating that of the third trimester of intrauterine life, without imposing stress on the developing metabolic and excretory system" (AAP 1977, Brady 1982). The overall goals for the nutritional management of premature infants include: supporting a rate of postnatal growth similar to intrauterine growth rates; promoting organ development necessary for extrauterine life; preventing nutrient deficiencies and avoiding nutrient excesses; and fostering optimal neurological development (AAP 1998,Anderson 2002, Wilson 1995). The nutritional management of premature infants can be categorized into four phases: acute, transitional, stable-growing and pre-discharge (Carlson 1999). Each phase consists of the appropriate nutrition support related to the infant's medical stability, maturation of metabolic and physiologic systems and developmental readiness (Anderson 1999, Fulham 2000, Jones Wessel 2000, 13 Literature Review Ziegler 2002). Each phase has unique feeding and nutrition characteristics, nutrition goals, nutrition requirements and targeted growth rates. The "acute phase" consists of the first 2 weeks after birth when infants are primarily fed parenteral nutrition and receive minimal enteral feeding (trophic feeds) to stimulate gut hormones and promote Gl maturation (Carlson 1999, Jones Wessel 2000). Maternal milk or preterm formula can be used for minimal enteral feeding (Jones Wessel 2000). The nutrition goal is to provide sufficient nutrients to prevent deficiencies and substrate catabolism. Premature infants may lose up to 15-20% of their birthweight in the first few days of life (Katrine 2000, Anderson 2002). The aim for growth during this phase is to minimize weight loss and regain birth weight. The "transitional phase" corresponds to the next 2-3 weeks of life, as the infant is transitioned from parenteral nutrition to enteral nutrition (Carlson 1999). Achieving full enteral feeds that supply adequate calories and nutrients to support growth at intrauterine rates is the nutritional goal. Types of feeds include fortified maternal milk (addition of a commercial multi-nutrient human milk fortifer) and premature infant formulas (Anderson 1999). The "stable-growing" phase describes infants who are more clinically and metabolically stable (Carlson 1999). Enteral nutrition is the primary source of nutrition support. During this phase infants are becoming more established on oral feeding, whether at the breast, cup and/or bottle. For infants who are not growing well, a variety of feeding/nutrition strategies may be employed to optimize intakes and growth. Current strategies to promote weight gain in infants receiving fortified human milk (expressed breast milk with a commercial human milk fortifier added) include: addition of infant formula powder to milk; addition of singular modules to milk (e.g. polycose powder, promod powder, MCT oil); increasing the infant's total fluid intake; or supplementing with a hypercaloric premature infant formula (Anderson 2002, Nevin-Folimo 2001, Sapsford 2000). The "pre-discharge" phase is a period when the infant is getting ready for home. It is a period of normalization of nutrition feeding regimes, discontinuation of premature products used in hospital (preterm formula and human milk fortifiers) and assessment of whether oral feeding on demand can sustain a satisfactory growth rate (Anderson 1999, Carlson 1999, Hovasi Cox 2000). 14 Literature Review 2.5 Enteral Nutrition and Premature Infants Enteral feedings are initiated in premature infants when their general condition allows it (Anderson 1999). Due to immature suck and swallow reflexes and medical conditions, during the first few weeks of life, many infants are tube fed into their stomach or small bowel (Jones Wessel 2000). Human milk is the recommended enteral feeding for all infants, including premature infants, because of its many nutritional, immunological and developmental benefits for the infant, as well as psychological benefits for the mother (AAP1997, Sapsford 2000, Schanler 1995). The potential short-term and long-term benefits associated with feeding breast milk to premature infants include: reduced incidence of infections (Narayanan, 1981, Narayanan 1982, El-Mohandes 1997), reduced incidence of necrotizing enterocolitis (Lucas 1990, Buesher 1994), improved feeding tolerance (Uraizee 1989), enhanced neurodevelopment (Lucas 1990), decreased number of hospitalizations (Meier 1996, Malloy 1993), enhanced family bonding, maternal involvement and interaction (Meier 1996, Kavanaugh 1997) and enhanced maternal self-esteem and maternal role attainment (Meier 1996, Kavanaugh 1997). Despite its many benefits, human milk alone does not meet the nutritional needs of the growing premature infant (Sapsford 2000). To better meet the energy, protein and vitamin/mineral needs of the premature infant, human milk is fortified with a commercial multi-nutrient human milk fortifier (Sapsford 2000, Schanler 2001). The Canadian Pediatric Society, C P S (1995) recommends the use of fortified preterm mother's milk or preterm infant formula (when mother's milk is unavailable) as the feeding of choice for premature infants (CPS 1995). For infants who are not receiving human milk, preterm infant formulas provide an alternative feeding option that is nutritionally adequate to support growth at intrauterine rates (Brady 1992, Rose 1993). These formulas are specifically designed to meet the needs of premature infants with increased nutrient content for growth and bone mineralization. Their nutritional composition is also designed to complement the capabilities of the immature infants metabolic and organ systems (especially Gl system). Term infant formulas are not appropriate for feeding premature infants during the neonatal period as they do not meet the specialized nutritional needs of the premature infant, in particular the protein, calcium and phosphorus requirements (Anderson 1999). For premature infants fed mother's milk fortified with a commercial human milk fortifier and who are growing poorly, having an alternative feeding strategy to increase 15 Literature Review the energy density of the milk without the use of exogenous supplements or additives is important for promoting the use of mother's milk and capitalizing on the many nutritional and non-nutritional factors that human milk provides (Kristen 1999). The hindmilk feeding strategy, which involves feeding premature infants a high fat energy dense milk fraction, has been shown to improve weight gain and energy intake of premature infants (Valentine 1994, Vasan 1998, Kristen 1999). 2.6 Milk Production by Mothers Who Deliver Premature Infants Mothers of premature infants wishing to breastfeed their infants at a later stage, or provide breast milk to their infant, must develop a good milk supply in the absence of a suckling infant. Artificial stimulation of the breast by mechanical expression is necessary for milk production. An electric pump is recommended for mothers who intend to pump regularly for an extended period of time (Hurst 1999). Mothers are required to invest a considerable amount of time and effort in expressing their milk as it may take many weeks to months before breastfeeding is established. For example, a mother who delivers a premature infant born at 28 weeks gestation, may pump for more than 8 to12 weeks before breastfeeding is successful. Early initiation of milk expression (within the first 6-24 hrs after delivery), frequent milk expressions per day (every 2-3 hrs in the first few weeks postpartum) and good breast emptying are reported to help establish an optimal milk production (Spicer 2001). The frequency of milk expression is a major factor in maintaining adequate milk production. DeCarvalho et al. (1985) reported higher milk volumes in mothers of premature infants who pumped > 4 times per day compared to those who pumped < 3 times per day (DeCarvalho 1985). Establishing and maintaining an adequate milk supply is a common challenge for mothers of premature infants (Groh-Wargo 1995, Schanler 1999). Milk volume production may fluctuate over the many weeks of pumping. A decrease in milk volume has been reported at approximately 4-6 weeks postpartum in mothers expressing their milk for prolonged periods (Hill 1996). Factors reported to affect milk production include: timing of initiation (Hurst 1999), pumping frequency and duration (De Carvalho 1985, Hopkinson 1988), type of pumping (manual versus electric) (Green 1982, Kubit 2000), method of pumping (double versus single) (Groh-Wargo 1995,Hill 1999,Hill 1996), smoking (Hopkinson 1992), alcohol intake (Cobo 1973,Howard 1999), and stress 16 Literature Review (Hurst 1997, Lau2001). Mothers may use a variety of techniques to improve their milk production, such as, infant contact (skin to skin cuddling) (Hurst 1997), relaxation/imagery techniques (Feher 1989), medications (motillium or metoclopramide) (Ehrenkranz 1986, Da Silva 1999), herbal galactagogues (e.g. fenugreek) (Hurst 1999, Howard 1999) and breast massage (Lau 2001). There is a wide variability in milk volume production among women (Groh-Wargo 1995, Hill 1999). Hopkinson et al. (1988) reported the milk production of 32 nonsmoking mothers of premature infants at 4 weeks postpartum (Hopkinson 1988). The mean (± SD) daily milk volume produced was 606 ± 369 mL (range 43-1773 mL). Other researchers have reported mean daily milk volumes of mothers who deliver premature infants ranging from 395 mL to 490 mL (Anderson 1981,Schanler 1980, Lemons 1982, Atkinson 1980). Recommendations for establishing and maintaining optimal milk production have been reported. Hopkinson et al. (1988) reported optimal milk production for mothers of premature infants occurs with at least 5 milk expressions per day and pumping durations of over 100 minutes per day (Hopkinson 1988). Target milk volumes for mothers of premature infants of 750-800 mL/day have been recommended to provide a safeguard (or reserve) in the event milk volume decreases during their infant's hospitalization (Hurst 1999, Schanler 1999, Spicer2001). Mothers who have established an adequate target milk volume may have an overabundant milk supply in the early weeks postpartum because their production of milk will be more than their infant's enteral intake. For example, a 1.5 kg premature infant enterally feeding at 150 mL/kg/day would be getting 225 mL/day. If the mother is producing 750 mL /day, she will have an excess of 525 ml/day or 3.3 times more milk than what her infant requires at that time. For mothers who produce an excess of milk relative to their infant's intake, collection of hindmilk (higher fat content breast milk) may be a consideration, especially if there are concerns with the infant's growth (Hurst 1999). 2.7 Fat Content of Human Milk Fat is the most variable macronutrient in human milk, representing approximately 40-50% of the total calories (Sapsford 2000). The fat content of milk varies with gestational age, stage of lactation, time of day, between women, and during a feed or pumping session (Jensen 1989a, Jensen 1989b). The average fat content of mature 17 Literature Review milk (>14 days postpartum) is 3.5 to 4.5 g/1 OOmL for women who deliver at term or prematurely (Hamosh 1994). 2.7.1 Inter-Individual Variation in the Fat Content of Milk A large variation in the fat content of milk occurs among women (Jensen 1989a, Hytten 1954e). Several investigators have demonstrated a wide range in milk fat concentrations, ranging from approximately 1.5 to 6 g fat /100mL (Anderson 1981, Clark 1982, Hytten 1954d). Hytten (1954d) provided comprehensive data on the individual fat content of milk from 150 mothers expressing milk on day 7 postpartum and from 30 mothers expressing milk after 20 days postpartum (mature milk). Subjects were considered representative of a random sample of patients from a maternity ward. Mothers expressed their milk 5 times per day using a mechanical pump to collect a 24 hour representative milk sample. The mean (± SD) fat content of milk on day 7 postpartum was 3.20 ± 0.81 g/1 OOmL, ranging from 1.60 to 5.9 g/1 OOmL (the mean and SD were calculated from the data published by Hytten). The mean (± SD) fat content of mature milk was 3.64 ± 0.79 g/1 OOmL, ranging from 1.45 to 4.9 g/1 OOmL. Table 2.2 summarizes selected studies that have reported the fat content of milk produced by mothers who have delivered prematurely (preterm milk) or at term (term milk). 2.7.2 Gestational Age and Variation in the Fat Content of Milk Several investigators have examined the influence of gestational age on the fat content of human milk. Anderson et al. (1981) reported a significant difference (P<0.05) in the milk produced by mothers delivering prematurely compared to mothers delivering at term. The fat content of preterm milk (n=12-15) was higher at all stages of lactation (day 3-5, day 8-11, day 15-18, day 26-29) compared to term milk (n=6-10). The investigators found that preterm milk was about 20-30% higher in fat concentration compared to term milk (Anderson 1981). LePage et al. (1984) compared the fat content of milk during the first month of lactation from mothers who delivered infants at 26-31 weeks gestation (very preterm milk, VPT, n=10), at 32-36 weeks gestation (preterm milk, PT, n=10) and at term (term milk, T, n=8). These investigators found that VPT milk had a higher (mean ± SE) concentration of fatty acids (4.46 ± 0.17 g/1 OOmL) than PT milk (3.94 ± 0.20 g/1 OOmL) 18 Literature Review Table 2.2 Selected References on the Fat Content of Mature Milk Produced by Mothers of Preterm (PT) and Term (FT) Infants1 Reference Gestational Age at Birth (wks) Method of Milk Collection Method of Fat Determination Stage of Lactation (days postpartum) Fat Content of Milk (g/100mL)2 (n) Anderson et al. (1981) PT = 26 -33 FT = 38 - 40 24 hr collection; manual / electric pump; 4-6 expressions/day; both breasts Solvent extraction With chloroform: methanol (2:1); measured colorimetrically Day 15-18 PT 4.33 + 0.24 (n=15) FT 3.06 ±0.21 (n=7) Atkinson et al. (1981) PT = 28.3 (mean) 24 hr collection Solvent extraction With chloroform: methanol (2:1); measured colorimetrically Day 13-15 4.05 ±0 .36 (n=8) Bitman et al. (1984) PT =31-36 V P T 3 =26-30 FT =37-40 Single morning collection (9-10 am); electric pump Folch method using 0 .01%BHT 4 ; measured gravimetrically Day 21 All 3.45 ± 0.37 (n=25) Butte et al. (1984) PT = 30 - 36 FT = 37-42 Single morning collection (between 8am to noon); electric pump Roese-Gottlieb Method 2 wks 4 wks PT 3.45 ± 0.94 b (n=8) FT3.79±1.00 5 (n=13) P T 3 . 9 5 ± 1.245(n=8) FT 4.04 ±0.78 5(n=13) Clark et al. (1982) Not reported Morning (between 9:30-11:30am) and afternoon (between 1:30-3:30pm) collection ; one breast; electric pump Modified Folch procedure; measured gravimetrically 2 wks 3.90 (n=10) Ehrenkranz et al. (1984) PT = 26-33 24 hr collections; 4-6 expressions/day; primarily electric pump (hand/hand pump also) Folch method; measured gravimetrically Day 14 Day 28 3.85 ±0 .54 (n=17) 3.78 ±0.32 (n=14) Gross et al. (1980) PT = 28-36 FT = 38-42 Single morning sample; mechanical/manual pump; both breasts Roese-Gottlieb Method Day 14 Day 21 Day 28 PT 4.40 ±0.31 (n=21) FT 3.48 +0.40 (n=12) PT 3.68 ±0.40 (n=15) FT 3.89 ±0.49 (n=12) PT 4.00 ±0.33 (n=13) FT 4.01 ±0.30 (n=11) Hytten (1954d) Not reported 24 hr collection; mechanical pump; 5 expressions/day Gerber method >Day 20 3.64 ± 0 . 7 9 b B (n=30) Lemons et al. (1982) PT = 27-37 FT = 39-41 24 hr collection; both breasts Frings and Dunn Method (colorimetric method based on the sulfo-phosoho-vanillin reaction) Day 14 Day 21 Day 28 PT 3.42 ±0 .18 (n=18) FT 3.00 ± 0.26 ( n=7) PT 3.52 ±0 .22 (n=17) FT 3.04+ 0.33 (n=7) PT 3.24 ± 0.16 (n=16) FT 3.07 ±0.21 (n=7) LePage et al. (1984) V P T J 26-31 PT 32-36 FT 24 hr collection; manual /mechanical pump Gas Chromatography 1 month VPT4.46±0.17 /(n=10) PT3.94± 0.20 7(n=10) FT 3.20 ± 0.30 7 (n=8) > — 1 \ 1 [_- 1 1— M a t u r e milk is defined as milk produced >14 days postpartum; M e a n ± S E M unless otherwise indicated; VPT=very preterm infant ; 4 B H T = butylated hydroxytoluene; 5 M e a n ± S D ; Calculated from published data; 7fatty acid content g / 1 0 0 m L 19 Literature Review or term milk (3.20 ± 0.30 g/1 OOmL). Very preterm milk had a significantly higher concentration of fatty acids (P<0.01) compared to term milk, but was not significantly different from PT milk. The fatty acid content of preterm milk and term milk were not significantly different. Other investigators have found little difference in the fat content of milk between preterm and term milk (Jensen 1989a). Lemons et al. (1982) compared the composition of milk from mothers delivering prematurely and at term during the first month of lactation. No significant difference was found between the preterm and term groups; the mean fat concentration for preterm milk was 3.28g% and for term milk was 3.02g% (Lemons 1982). Gross et al. (1980) studied the composition of human milk produced by mothers giving birth prematurely (28-36 weeks) and at term. No significant differences were found in the fat content of milk between the two groups (preterm milk versus term milk) during the first month postpartum (Gross 1980). The mean (± SEM) fat content of milk at Day 28 for preterm milk was 4.00 + 0.33 g/1 OOmL and for term milk was 4.01 ± 0.30 g/1 OOmL (Gross 1980). Bitman et al. (1983) compared the fat content of milk from mothers of term and preterm infants during the first few weeks postpartum. The investigators examined milk produced by mothers of very preterm infants (VPT milk, 26-30 weeks gestation), preterm infants (PT milk, 31-36 weeks gestation) and term infants (T milk, 37-40 weeks gestation). They reported little difference in the fat content of milk among the maternal groups (Bitman 1983). During the first month of lactation (Day 3, Day 7, Day 21) the fat content of milk appeared similar for all three maternal groups, with term milk containing slightly higher fat concentrations at all three stages (Bitman 1983). Butte et al. (1984) reported data on the fat content of milk produced by mothers of preterm and term infants from week 2 to 12 weeks postpartum. The investigators found no statistical differences in the fat content of milk between preterm milk and term milk. The overall mean (± SD) fat content of preterm milk (n=8) was 3.92 ± 1.25 g/1 OOmL and term milk (n=13) was 4.31 ± 1 . 1 5 g/1 OOmL (Butte 1984). Although the variation in the fat content of milk between mothers of premature infants and term infants will not be relevant to the present study (as only milk from mothers who delivered premature infants will be collected and analyzed), this variation in the fat content is important to consider when reviewing the literature and interpreting results on the fat content of mother's milk. LePage et al. (1984) reported that mothers (n=10) who delivered premature infants born at a gestational age of 26-31 weeks had a 20 Literature Review higher fatty acid content of milk than mothers (n=10) who delivered premature infants born at 32-36 weeks gestation. This data was not significantly different (as discussed earlier), however, further studies (with a larger sample size) to examine potential differences in the fat content of milk among mothers who deliver premature infants at different gestational ages may be indicated. 2.7.3 Stage of Lactation and Variation in the Fat Content of Milk The fat content of human milk changes with the stage of lactation, from colostrum (early milk) to mature milk (Jensen 1989b). There are three stages of lactation: colostrum (early milk) refers to milk produced from day 1 to 5 postpartum, transitional milk refers to milk produced from day 6 to 14 postpartum and mature milk refers to milk produced after 14 days postpartum (Jensen 1989b, Lawrence 1999). The time periods postpartum that define the stages are not absolute and can vary within a few days (Jensen 1989b). Several investigators have reported an increase in the fat content of milk from mothers who deliver at term or prematurely during the first month postpartum (Anderson 1981, Atkinson 1981, Ehrenkranz 1994, Harzer 1983,Lemons 1982). Anderson et al. (1981) reported the fat content of milk produced by mothers of preterm and term infants over the first 4 weeks of lactation. The fat content of milk increased from the first week of lactation (Day 3-5) to the second week of lactation (day 8-11), and then remained stable (day 15-18; day 26-29) in both groups (Anderson 1981). The fat content of early preterm milk (day 3-5, n=12) was 3.10 ± 0.23 g/100 mL (mean ± SEM) and the fat content of mature preterm milk (day 15-18, n=15) was 4.33 ± 0.24 g/100 mL. The fat content of early term milk (n=10) was 1.85 ± 0.35 g/1 OOmL and for mature term milk (n=9) was 3.06 ± 0.21 g/100 m L Ehrenkranz et al. (1999) reported the total fat content of milk produced by mothers (n=21) who delivered premature infants during the first six weeks of lactation. The total fat content of milk increased significantly (P<0.001) from 1.99 ± 0.25 g/1 OOmL (mean ± SEM) in colostrum (day 2 postpartum) to 3.89 ± 0.25 g/1 OOmL in mature milk (>day 14 postpartum). Gross et al. (1980) reported an increase in the fat content of milk from mothers of term and preterm infants. The fat content of milk produced at day 3 postpartum by mothers of preterm and term infants was 1.63 ± 0.23.g/1 OOmL and 1.71 ± 21 Literature Review 0.24 g/100mL (mean ± SEM), respectively. The fat content of mature milk produced at day 28 by mothers of preterm and term infants was 4.0 ± 0.33g/100mL and 4.01 ± 0.30 g/100mL, respectively. The fat content of human milk remains relatively constant with prolonged lactation (Jensen 1989a). Several investigators have observed a relative constant concentration of fat in mature milk during the first 12 weeks of lactation produced by mothers who deliver prematurely (Bitman 1993, Butte 1984, Lemons 1982). Butte et al. (1984) studied the fat content of milk collected for 12 weeks postpartum by mothers (n=8) who delivered premature infants. The fat content of preterm milk did not change significantly over time and was reported to be 3.95 ± 1.24 g fat/1 OOmL (mean ± SD) at 4 weeks postpartum, 4.12 ± 1.28 at 8 weeks postpartum, and 3.82 ± 1.45 g fat/1 OOmL at 12 weeks postpartum (Butte 1984). Lemons et al. (1982) studied the fat content of milk collected by mothers of premature infants during the first several weeks of lactation. The fat content of mature preterm milk was also relatively constant over time and was reported to be (mean ± SE) 3.42 ± 0.18 g/1 OOmL (n=18) at 2 weeks postpartum, 3.24 ± 0.16 g fat/100mL(n=16) at 4 weeks postpartum, 3.43 ± 0.16 g fat/1 OOmL (n=11) at 6 weeks postpartum, and 3.46 ± 0.32 g/1 OOmL (n=9) at greater than 8 weeks postpartum (Lemons 1982). 2.7.4 Throughout the Day (Diurnal) Variation in the Fat Content of Milk Several investigators have reported diurnal variations in the fat content of human milk (Hall 1979, Harzer 1983, Hytten 1954c). Hytten (1954c) described a pattern of varying levels of fat concentrations in human milk over the course of the day. He examined milk expressed by women (n=42) at 5 specific time periods in the day (6 am, 10 am, 2 pm, 6 pm and 10 pm). The fat content of mature milk was lowest at 6 am (2.64 g /1 OOmL), peaked at 10 am (4.15 g /1 OOmL) and then gradually decreased during the following 3 time periods (3.81g/100mL at 2pm; 3.58 g/100mL at 6pm and 3.12 g/100mL at 10 pm) (Hytten 1954c). Hall (1979) described a pattern of diurnal variation in the fat content of milk of one mother at 4 time periods (6 am, 10 am, 2pm and 6 pm). The mean fat content of milk was lowest at the early morning period (6 am) and increased during the day to the highest level at 2 pm, thereafter the fat content appeared to plateau (Hall 1979). The fat content of milk increased about 2.5-fold from lowest value in 22 Literature Review the early morning to the highest value at midday (Hall 1979). Due to the known daily variation in the fat content of mother's milk, it is recommended to collect a complete 24 hour milk sample for lipid analysis (Jensen 1989a). 2.7.5 Within a Feed Variation in the Fat Content of Milk It is well documented that the fat content of human milk changes within a feed (or pumping session). Many investigators have observed an increase in the fat content of milk during a nursing (or pumping) session (Dorea 1982, Emery 1978, Forsum 1979, Hall 1979, Hytten 1954b, Macy 1931, Nelville 1984, Prentice 1981, Valentine 1994). The magnitude of the increase in fat content from the beginning of a feed to the end of a feed may vary due to individual differences among women and/or differences in the method of collecting, sampling and analyzing milk (Macy 1931, Jensen 1989a). The fat content of milk during a feed (or pumping session) has been reported to increase approximately 1.7 - to 3- fold from foremilk to hindmilk (Nelville 1984,Valentine 1994). Table 2.3 summarizes selected studies that have reported a within feed variation in the fat content of human milk. Hytten (1954b) provided comprehensive data on the within feed variation in human milk composition, by analyzing the fat, nitrogen and lactose concentrations in mothers' expressed milk. Twenty mothers collected 12 mL consecutive samples during a pumping session. Hytten (1954b) demonstrated a progressive increase in the fat concentration of milk during the course of the pumping session. The fat content of milk in the first 12 mL fraction ranged from 0.7-5.6 g/1 OOmL and the fat content of milk in the last 12 mL fractions ranged from 2.5-10.6g fat/1 OOmL. No statistical procedures were conducted on the data; however, a common trend in the change in fat content of the milk was observed during the pumping sessions. In general, the fat content of milk was low at the beginning of the pumping session, increased progressively during the session, and had a rapid rise in fat content at the end of the pumping session (Hytten 1954b). Forsum and Lonnderdal (1979) studied 4 mothers who collected 2-10 mL sequential milk fractions during one nursing period in the day. The authors reported a similar trend as Hytten in the fat content of milk during a nursing session. The increase in the fat content of milk observed during the first half of the nursing period was associated with the milk ejection reflex. Another increase in milk fat content was 23 Literature Review Table 2.3 Selected References Summarizing the Within Feed Variation in the Fat Content of Human Milk Reference Subject / Characteristics Method of Milk Collection Method of Lipid Analysis Results of Study Macy et al. (1931) 3 mothers Manual expression; milk separated into first half and last half collections of a feed; 24 hour collection Micro -Babcock Method Range of fat content of milk: First half: 3.26- 5.60 g/1 OOmL Last half: 4.75- 6.56 g/1 OOmL Milk fat T during a feed. Hytten (1954 b) 20 mothers (53 milk samples) Mechanical expression; Sequential 12 mL collections of milk during a feed Gerber Method No statistical data provided. Range of fat content of milk: First 12mL fraction: 0.7-5.6g /100mL Last 12 mL fraction: 2.5 -10.6g/1 OOmL Milk fat T during a feed. Emery et al. (1978) 3 mothers Manual expression; 12 mL sequential milk fractions during a feed; complete expressions Erickson & Dunkley Method; measured gravimetrically Results reported graphically. The % of fat in the milk fractions increased during a single nursing period. Hall (1979) 6 mothers (15 milk samples) 5-42 wks pp 1 Hand expression; Sample of milk collected before and after a breastfeeding session Folch Method Fat content of Milk (mean±SE) Before Feed: 2.42 ±0 .3 g/1 OOmL After Feed: 7.48 ±0.57 g/1 OOmL Milk fat T during a feed. Forsum & Lonnderal (1979) 4 mothers 1-6 months pp 1 Electric pump; 5 mL milk sample before and after a breastfeeding session Zollner and Kirsh Method; (commercial kit) Results reported graphically. Milk fat T during a feed. Prentice et al. (1981) 60 mothers 1-18 mo pp 1 Gambia "Maternal" expression; small milk sample taken before and after a breastfeeding session Creamatocrit Method 2 Results reported graphically. Milk fat t during a feed. Dorea e ta l . (1982) 17 mothers 15 -60 days pp 1 Brazil (82-87 milk samples) Manual expression; 5 mL milk sample taken before (foremilk) and after (hindmilk) a breastfeeding session Roese-Gottlieb Method Fat content of Milk (mean±SD) Foremilk: 2.33 ± 0.81 g/1 OOmL Hindmilk: 4.22 ± 1.96 g/1 OOmL Milk fat T during a feed. Nelville et al. (1984) 10 mothers 33-210 day pp 1 Manual expression; 5 mL milk sample before (foremilk), mid-way, and after (hindmilk) a breastfeeding session Creamatocrit Method 3 Fat content of Milk (mean) Foremilk: 2.3 % Hindmilk: 4.6 % Milk fat T during a feed. Valentine eta l . (1994) 15 mothers delivering preterm infants Electric pump; 24 hr collection; foremilk designated as milk collected 2-3 minutes after letdown; hindmilk designated as the remainder of milk collected during a pumping session Gravimetric Method (Jeejeeboy) Fat content of Milk (mean±SD) Foremilk: 2.86 ± 0.81 g/1 OOmL Hindmilk: 4.78 ± 0.85 g/1 OOmL Milk fat t during a feed. pp refers to days postpartum (stage of lactation) 2creamatocrit values (%) calibrated to fat values (g/1 OOmL) determined by a modified British Standard Method 3creamatocrit values (%) calibrated to fat values (g/1 OOmL) determined by the modified Folch Method 24 Literature Review observed during the last few millilitres of milk volume expressed at the end of the session. The investigators reported there was considerable variation between subjects in the range of fat content in milk during a feed. Dorea et al. (1982) studied 17 mothers in Brazil, who manually collected a 5 mL sample of milk before (foremilk) and after (hindmilk) a breastfeeding session. The results showed a significantly higher concentration of fat in hindmilk than foremilk (P<0.001). The mean (± SD) fat content of foremilk and hindmilk was 2.33 ± 1.50 g/100 mL and 4.22 ± 1.96g/100mL, respectively. Hall (1979) reported data on 6 mothers who collected milk samples before and after a breastfeeding session. The fat content of milk increased 3-fold from foremilk to hindmilk. The mean (± SE) fat content of foremilk and hindmilk was 2.32 ± 0.30 g and 7.48 ± 0.57g/100mL, respectively. No statistical data were provided, however, the researcher reported there was a "significant" increase in lipid concentration during the feed. Neville et al. (1984) studied 10 breastfeeding mothers of term infants who collected a 5 mL milk sample at the beginning of a feed (foremilk), in the middle of the feed (mid-milk) and at the end of a feed (hindmilk). Neville et al. (1984) reported a significant difference in the fat concentration between foremilk and hindmilk (P=0.001). The fat concentration of foremilk was 2.3% and hindmilk was 4.6%, a 2-fold increase in lipid concentration from foremilk to hindmilk. Despite differences in the methodology of the studies, in particular the sample size, method for collecting the milk samples, and lipid analysis procedures, all the studies showed a higher fat content in hindmilk compared to foremilk. 2.8 Hindmilk Feeding Over the past few years there has been an interest in the use of hindmilk to improve weight gain in premature infants. Published research on fractionating maternal milk into foremilk and hindmilk for the purpose of feeding hindmilk to premature infants is limited. One study by Valentine et al. (1994) and abstracts by Slusher et al. (2000) and Vasan et al. (1998) have examined hindmilk feeding as a strategy to promote weight gain in premature infants. Valentine et al. (1994) in a non-randomized prospective study examined the effect of feeding hindmilk (fortified with HMF) to 15 low birth weight infants (mean birth weight 1087g, mean gestational age 28 wks) to improve weight gain. The method of separating milk into 2 fractions, foremilk and hindmilk, was based on a predetermined time period. 25 Literature Review The first milk fraction, foremilk, was collected until 2-3 minutes after milk flow began ("letdown") and the later milk fraction, hindmilk, consisted of milk collected thereafter. Milk samples were collected for a 24 hour period. Infants were followed during two study periods: the baseline period (Week 1) and the hindmilk feeding period (Week 2). The infants' enteral intake was similar for both study periods, approximately 158-159 mL/kg/day. The infant's energy intake was higher during the Week 2 (132 ± 16 kcal/kg, mean ± SD) compared to Week 1 (117 ± 18 kcal/kg). The infants' rate of weight gain was significantly higher during Week 2 (rate of weight gain ranged from 12-21 g/kg/day) compared to Week 1(rate of weight ranged from -0.7 to 13 g/kg/day), with an average increase of 7.0 ± 4.4 g/kg/day. The mean fat content of hindmilk was significantly higher than the fat content of foremilk, 4.78 ± 0.85 g fat/1 OOmL and 2.86 ± 0.81 g fat/1 OOmL, respectively. The higher content of fat in hindmilk, corresponded to a higher energy content in hindmilk. The mean energy content of hindmilk was 82.4 ± 7.7 kcal/100mL and foremilk was 62.9 ± 10.8 kcal/100mL. The investigators concluded that feeding hindmilk, higher fat content milk, is a simple strategy to increase the rate of weight gain in low birth weight infants. However, the investigators did randomize the order of feeding, which would have strengthened this conclusion. Slusher et al. (2000) examined the effect of feeding hindmilk to 18 Nigerian low birth weight infants (mean birthweight 1385g, range 904-1718g). The method of separating milk into foremilk and hindmilk was not clearly defined, however, the investigators reported that the creamatocrit procedure was used to determine the point during milk collection when separation occurred. Hindmilk was fed to the infants for 10 days, after which the infants returned to composite (foremilk + hindmilk) milk feedings until discharge. The mean creamatocrit value for foremilk was 5% (range 2.5-9%) and for hindmilk was 8.5% (range 6-11%). The mean foremilk volume collected was 199 mL/day (range 110-387mL) and the mean hindmilk volume collected was 366mL/day (range 90-170 mL). The investigators reported the mean daily weight gain during the hindmilk feeding period was 19.7 g/day (range 12-24 g/day), which appears to fall within the targeted intrauterine daily (average) growth rate of 11-25 g/day for infants with a birth weight of 900 to 1700g (Katrine 2000). Vasan et al. (1998) in a prospective study examined the effect of individualizing the lipid content of mother's milk on weight gain in extremely low birth weight infants. 26 Literature Review Nineteen of the 21 infants followed in the study required hindmilk feedings due to poor weight gain (<15g/day) while on fortified composite milk (foremilk + hindmilk). Two infants had acceptable weight gains (>15g/day) on fortified composite milk feedings and therefore did not require hindmilk feedings; the creamatocrit value for composite milk fed to these infants was 10-13%. Nineteen infants in the study received hindmilk or hind-hindmilk (fractionating later in the milk expression procedure) for a 10 day study period. The mean creamatocrit value for hindmilk was 11.3% (range 9-18%) and the mean creamatocrit value for hind-hindmilk was 13.3% (range 11-15%). The mean rate of weight gain during the hindmilk/hind-hindmilk feeding period for the 19 infants was 19.7 g/day. This rate of weight gain was significantly higher than the rate of weight gain (3.8g/day) during the composite feeding period for the same infants. The mean creamatocrit value for composite milk (n=19) was 8.4% (range 6-10%). 2.9 Fractionating Human Milk into Foremilk and Hindmilk There is no standard method for separating expressed human milk into foremilk and hindmilk. There are three possible methods to fractionate milk into foremilk and hindmilk: the time method, the volume method and the visual method. The time method involves separating milk based on a designated time from "letdown" or the start of pumping. The volume method involves separating milk based on a designated volume or proportion of milk expressed during a pumping session. The visual method involves separating milk based on the visual appearance of the expressed milk. The fractionating method used by Valentine et al. (1994) was based on a designated time period for determining the point during the milk expression session when milk was separated into of foremilk and hindmilk. The investigators defined foremilk as the milk collected for 2-3 minutes after "letdown" (increased milk flow) and hindmilk as milk expressed thereafter (Valentine 1994). This is a simple method for fractionating milk as it requires minimal equipment and appears easy to perform. Valentine et al. (1994) found a 1.7 fold difference in fat content from foremilk to hindmilk using this method (Valentine 1994). Lang (1997) describes a method of separation based on a set volume or proportion of milk expressed during each pumping session (Lang 1997). This fractionating method takes into consideration the mother's milk volume relative to their infant's daily milk intake. For example, if the mother's daily expressed volume is 300 mL and the infant's 27 Literature Review daily intake is 200 mL, there would be an excess of 100 mL milk per day or 33% more milk than what is needed to meet the infant's prescribe enteral intake. The point of separation would be when the mother has expressed 33% of the volume typically pumped at that session. This first milk fraction collected is designated as foremilk. This method has not been reported in any published studies, but has been recommended by Lang for the promotion of weight gain in preterm infants (if the mother has a good milk supply). This method requires more work on the mother's part for recording volumes and calculating the proportion of milk designated as foremilk, but it may provide more accurate results. Another method for separating milk is by visual observation. The point of separation is based on a change in color and consistency of breast milk from a thin, white colored milk (foremilk) to a thick, creamier milk (hindmilk). The point of separation is subjective and arbitrary and may affect the accuracy of this method. To date, these three fractionating methods have not been compared in terms of their practicality (mothers' evaluation of practicality, overall preference, and willingness to incorporate the method into their usual milk expression routine), efficacy (change in fat content from foremilk to hindmilk and adequacy of hindmilk volume to meet the infant's enteral intake) and reliability (can the method be reproduced over time). 2.10 Creamatocrit Measurements and Clinical Practice The Creamatocrit Method is a simple and quick technique first described by Lucas et al. (1978) for the estimating the fat content and energy content of human milk (Lucas 1978). The method has been widely used in both clinical and research settings. The creamatocrit method has been validated by several researchers and felt to be a reliable technique for measuring the fat concentration of human milk (Lucas 1978, Lemons 1980,Meier 1999, Wang 1999). This method involves centrifuging a small amount of milk (< 1 mL) and measuring the cream layer that has been separated during spinning. The "creamatocrit" percent value is the height of the cream layer relative to the height of the total milk column. Several researchers have found a strong positive linear . relationship between the creamatocrit percent value and the milk fat content quantified by standard methods of fat analyses (Lucas 1978, Lemons 1980, Meier 1999, Wang 1999). The estimated fat and energy content of human milk can be calculated by using published regression equations (Lucas 1978,Lemons 1980,Meier 1999). 28 Literature Review In clinical practice, infants fed human milk are assumed to be receiving milk containing 4 g fat/1 OOmL and 70kcal/100mL. The large variation in the fat content of milk among women suggests that some infants will receive milk with a lower fat content and subsequently lower energy concentration than expected. Infants receiving low fat, low energy dense milk are likely to be at risk of suboptimal energy intake and poor growth. Knowledge of the nutrient composition, in particular the fat and energy content of milk, would assist clinicians in: (1) calculating a more realistic estimation of the energy intake of the infant; (2) identifying a mother who produces a milk with a low fat content (and hence lower energy content), which may place their infant at risk of receiving a suboptimal energy intake; or (3) individualizing the infant's enteral feed with respect to fortification, supplementation, or implementation of hindmilk feedings. Based on these reasons, the Creamatocrit Method could be a useful clinical tool in the nutritional management of premature infants fed maternal milk. In one NICU in the USA, mothers are using the creamatocrit method to facilitate lactoengineering, separating milk into foremilk and hindmilk, for hindmilk feeding (Grifffin 2000). In a research setting, the advantages of using the creamatocrit method over more standard quantitative methods are: it is less expensive; it takes less time; it is simple to perform; and does not require exposure to chemicals/solvents or expensive laboratory equipment (Jensen 1989a, Lucas 1978,Polberger 1993). 2.11 Limitations To Our Present Knowledge Currently, premature infants are fed enteral feeds of expressed mother's milk fortified with added protein, carbohydrate, vitamins and minerals on a volume basis without the knowledge of the nutrient content of the milk. Infants are assumed to be receiving a set amount of calories based on the average fat content of the milk (3.5-4.5 g/1 OOmL). A preliminary study (unpublished) at B C C H of premature infants fed preterm human milk, found the variation of milk fat content between women to be 1.8 to 4.7 g fat/100 mL. The estimated energy content of these mothers' milk would range from 51-76 kcal/100mL, a difference of 26 kcal/100mL. Some infants would be at risk of receiving suboptimal calories necessary for optimal weight gain and growth due to the low energy density of their mother's milk. A practical feeding strategy of fractionating human milk into foremilk and hindmilk, and feeding fortified hindmilk may improve weight gain in these infants. 29 Literature Review Mothers who deliver premature infants may express their milk for many weeks to months before breastfeeding is established. Some mothers may produce an abundant supply of milk due to the artificial stimulation of their breasts by mechanical pumping. For mothers who express milk volumes in excess of their infant's prescribed intake, fractionating milk into foremilk and hindmilk, with subsequent feeding of the energy dense hindmilk fraction to promote weight gain, may be an option. At present, there are no standardized methods for separating human milk into two fractions: low energy dense milk and high energy dense milk. No studies have been conducted that examine the practicality, accuracy or reliability of potential methods for fractionating milk for the selective collection of hindmilk. Determining a fractionating method that is practical and accurate will allow subsequent research to examine the effects of feeding high energy dense milk fractions to premature infants. The development of a fractionating protocol for the selective collection of hindmilk could be used in clinical practice as part of a feeding strategy to promote weight gain in premature infants. The hindmilk feeding strategy is an alternative method for increasing the energy intake of premature infants without the use of exogenous supplements and additives such as, term infant formula powder, MCT oil, polycose powder, or hypercaloric preterm infant formula. This feeding strategy supports the lactation efforts of mothers and utilizes the nutritional and non-nutritional benefits of feeding human milk to premature infants. Meier (1998) reported finding an unexpected outcome from her studies on hindmilk feeding. She found that mothers became "invested" in providing hindmilk to their infants and that this contribution to their infant's care may serve as a powerful motivator to sustain milk expression during their infant's hospitalization (Meier 1998). 30 Design and Methods CHAPTER 3 DESIGN AND METHODS 3.1 STUDY DESIGN This study was conducted in three sequential phases: Phase I - Development of a fractionating protocol for the collection of hindmilk; Phase II -Reproducibility of a fractionating protocol for the collection of hindmilk; and Phase III -Examination of breast milk expression practices, expressed breast milk volumes and factors that may affect breast milk supply. This study involved a convenience sample of mothers whose premature infants were patients in the Special Care Nursery of B.C.'s Children's Hospital. The following chapter describes the design and methodology used in each of the three study phases. The first sections of this chapter describe the development of the nutrition study booklet, the recruitment procedure, and the participant selection criteria. The next section describes the study procedures for each of the three study phases. The last two sections of this chapter describe the laboratory methods used in this study and the statistical tests used to analyze the study data. 3.2 DEVELOPMENT OF NUTRITION STUDY BOOKLET A nutrition study booklet was developed to collect study information. The booklet for Phase I and Phase II described the study procedures and consisted of 4 sections: (1) a milk expression diary (data collection table) for the baseline period and each of the three fractionating periods; (2) an expressed breast milk practices survey; (3) an evaluation for the fractionating methods; and (4) a socio-demographic questionnaire. Mothers were requested to provide feedback or comments regarding the study procedures or breast milk expression observations in the booklet. The nutrition study booklet for Phase III consisted of a 3-day milk expression diary, an expressed breast milk practices survey, and a socio-demographic questionnaire. Several steps were undertaken to support face validity and content validity during the development of the milk expression diary and data recording instruments. Protocols for the 3 fractionating methods were developed from information obtained from the literature and clinical experience. The protocol for each of the 3 methods was pilot-tested with a group of mothers (n=5) in the SCN who had delivered premature infants before incorporation into the study booklet. The survey on breast milk expression practices and factors that may affect milk production was developed based on 31 Design and Methods information obtained from the literature, experts in the field and clinical experience. Individuals with an expertise in nutrition, human lactation, neonatology and survey design were consulted to review sections of the nutrition study booklet and offer comments for improvements in content, clarity and format. The study booklet was then pre-tested with a group representative of the target population 3 mothers in the S C N who delivered premature infants). Feedback and suggestions from the mothers and experts were incorporated into the final booklet before using it in the study. 3.2.1 Milk Expression Diary Mothers were asked to record information on their milk pumping routine in the milk expression diary for the baseline period and fractionating periods. The baseline period (the first 24 hour study period) was conducted at the beginning of the study to collect baseline information about the mothers' usual milk expression routine and milk production. The fractionating periods consisted of 3 separate study days, 1 study day for each fractionating method. Data were collected on the mothers' milk expression routine and milk production during each of the 24 hour fractionating periods. The milk diary provided information on time of day, duration (total minutes pumping per day), location, type of pump used (electric/hand), frequency (number of pumping sessions per day), method of pumping (single or double), side of breast pumped (right and/or left) and volume of milk expressed at each milk expression session. Duration of pumping was calculated as the sum of expression time for both breasts within a pumping session. The single pumping method consisted of mothers pumping 1 breast, followed by the second breast during a single pumping session (sequential pumping). The double pumping method consisted of mothers pumping both breasts at the same time during a single pumping session (simultaneous pumping). For mothers who double pumped, the pumping time was multiplied by 2 to obtain the total minutes pumped during that session. 3.2.2 Expressed Breast Milk Practices Survey A survey was developed to provide exploratory information on the practices of breast milk expression of mothers who deliver premature infants, and to assess the prevalence of factors that may affect breast milk volume (e.g. letdown, stress, alcohol intake, cigarette smoking, medications, and herbal products). The survey consisted of 32 Design and Methods both open and closed-ended questions. The survey was conducted in each of the three study phases. 3.2.3 Evaluation Mothers evaluated the three fractionating methods subjectively in Phase I. The preferred fractionating method was determined based on a practicality score, ranking of the methods and willingness to perform the method for an extended period of time. The practicality of the fractionating methods was determined by comparing the methods on the following criteria using a Likert Scale (rating scale of 1 to 5) for three categories, with 5 being the highest value: (1) time to complete the fractionating method as compared to the usual expression time; (2) clear and understandable directions; and (3) convenience of the method. A practicality score was calculated for each of the three methods based on the sum of the ratings for the three criteria. The maximum practicality score for each method was 15. Mothers were also asked to rank (1 to 3) the fractionating methods for overall preference, where the most preferred method was designated a value of 1 and the least preferred method was designated a value of 3. Mothers were asked if they were willing to perform the method for an extended time period. 3.2.4 Socio- Demographic Questionnaire A questionnaire was developed to collect socio-demographic characteristics about the study participants, such as, age, country of birth, number of years resident in Canada, city of residence, marital status, highest level of education completed, usual occupation, gross family income per year, number of people dependent on the family income, ethnic background, number of children living in the household and ages of children living in household. Infant characteristics, such as date of birth, gestational age at birth, birth weight, birth length, birth head circumference, birth classification and weight, and gestational age of the mother's preterm infant at study entry were obtained from the infant's medical chart. 33 Design and Methods 3.3 STUDY PARTICIPANTS 3.3.1 Participant Recruitment Mothers who delivered premature infants were recruited between August 1999 and December 2001 from the Special Care Nursery, SCN, at British Columbia's Children's Hospital. Participants were recruited by posting a recruitment letter in the mothers' "pump room" located in the SCN (Appendix 1). A recruitment letter was also left at the bedside of infants for mothers who were considered eligible for this study (Appendix 2). Eligibility was based on the criteria in section 3.3.2. Each potential participant was then approached to ascertain their interest for possible enrollment in the study and verify their eligibility. The study procedures were described to each mother and informed written consent was obtained for interested participants who met the eligibility criteria (Appendix 3). 3.3.2 Participant Selection Criteria Eligible study participants for Phase I and II consisted of mothers who delivered premature infants (less than 37 weeks gestation), expressed a minimum of 120 mL per day of mature breast milk (greater than 14 days postpartum), expressed more than 125% of their infant's prescribed enteral fluid intake, and whose infant was receiving expressed breast milk as their sole source of enteral nutrition. Mothers were considered ineligible if they had a known history of substance abuse (illicit drugs and/or alcohol), HIV, hepatitis or other communicable diseases, or were unable to understand the informed consent and requirements for the completion of the study. Eligible study participants for Phase III consisted of mothers who delivered premature infants (less than 37 weeks gestation), expressed mature breast milk (greater than 14 days postpartum), and were able to understand the informed consent and requirements for the completion of the study. 3.4 ETHICS The study protocol and procedures were approved by the University of British Columbia Clinical Research Ethics Board and the British Columbia's Children's Hospital Research Review Committee. 34 Design and Methods 3.5 STUDY PROCEDURE 3.5.1 Phase I: Development of a Protocol for Fractionating Expressed Human Milk into Foremilk and Hindmilk The study procedures were explained to the mothers and a nutrition study kit, containing the nutrition study booklet (milk expression diary, survey, evaluation and socio-demographic questionnaire) (Appendix 4), syringes, labels, pens and stopwatch were given to the mothers. The mothers were instructed to record information and collect expressed breast milk for the baseline period and 3 fractionating periods. The length of time for each study period was 24 hours. The following diagram illustrates the study procedure for collecting milk samples and recording information. Baseline Period Fractionating Periods Practice Day Fractionating Method # 1 Practice Day Fractionating Method # 2 Practice Day Fractionating Method # 3 Milk Collection (24 h) •/ - •/ -Milk Expression Diary (24 h) - - -The first 24 h period represented the baseline period and the following 3 x 24 h periods represented the fractionating periods (1 study day for each fractionating method). The mothers were instructed on how to fractionate their milk for each of the 3 methods. The order of performing the fractionating method was randomly assigned using a sealed envelope containing 3 pieces of paper corresponding to each method. A practice day preceded each fractionating method study period. This was done to enable mothers to practice the procedure for separating milk into foremilk and hindmilk, and ask questions or clarify procedures if needed. The minimum number of days to complete the study requirements was 7 days. The mothers followed standard SCN guidelines for set-up and handling and storage of expressed breast milk, unless otherwise indicated in the study procedures. Milk samples were collected for the baseline period and for each of the 3 fractionating periods. The mothers were instructed to collect a 1-2 mL sample from every container (gradufeed) of milk expressed within each 24 h study period, using the syringes provided. Each participant's milk samples were pooled for each study period to represent a 24 h milk collection. The mothers were instructed on how to label samples using the labels provided. If mothers were unable or unwilling to collect the syringe sample, the nutrition 35 Design and Methods researcher col lected the milk sample . S a m p l e s were stored in the refrigerator on the unit, or at home and then brought to the S C N when the mother next visited the hospital. The pooled milk samp les were stored at -80°C until analys is. E a c h milk sample was ana lyzed in triplicate by the modif ied Folch method (Folch 1957) for the total fat content and the Creamatocr i t method (Lucas 1978) for the creamtocrit value. The basel ine milk samp les were a lso ana lyzed for total fatty acid content and medium chain fatty ac id content by direct methylation and gas-l iquid chromatography (GLC) . The mothers were asked to evaluate the 3 fractionating methods in order to determine the mothers' preferred fractionating method. The mothers were also asked to complete a survey on their pract ices of breast milk express ion and soc io-demographic background. Infant characterist ics were obtained from each infant's medical chart. 3.5.2 Phase II: Reproducibility of a Protocol for Fractionating Expressed Human Milk into Foremilk and Hindmilk The procedures for P h a s e II were similar to P h a s e I in that mothers were instructed to record information and collect expressed breast milk samp les for the 24 h basel ine period, then the 3 x 24 h fractionating periods. However, only 1 method for fractionating the expressed milk was performed during the fractionating study periods. Th is method was the preferred fractionating method determined in P h a s e I. The mothers were instructed on how to separate their milk into foremilk and hindmilk using the preferred fractionating method. The mothers were asked to perform the method for 3 x 24 h periods on non-consecut ive days (2 weekdays and 1 weekend day) to allow for any dif ferences in milk vo lume or pumping routine. A minimum of 2 fractionating col lect ion days was cons idered acceptable for study complet ion. The following diagram illustrates the study procedure for col lecting milk samp les and recording information in P h a s e II. Baseline Period Fractionating Periods Using the Preferred Fractionating Method Practice Day Fractionating Day #1 Day Off Fractionating Day #2 Day Off Fractionating Day # 3 Milk Collection (24 h) - -/ - -Milk Expression Diary (24 h) •/ - -36 Design and Methods A nutrition study kit, including the nutrition study booklet (Appendix 5), was given to each mother. Collection of milk samples and lipid analysis was similar to Phase I. The mothers were asked to complete a survey on their practices of breast milk expression and socio-demographic background. Infant characteristics were obtained from each infant's medical chart. 3.5.3 Phase III: Determination of Expressed Breast Milk Practices The study procedures were explained to the mothers and a nutrition study booklet (milk expression diary, expressed breast milk practices survey and socio-demographic questionnaire) was given to each mother (Appendix 6). Each mother was instructed to record information on the volume of milk expressed and their milk expression routine in the milk diary for 3 non-consecutive 24 hour study periods (2 weekdays and 1 weekend day). No milk samples were collected. The mothers were asked to complete a survey on their practices of breast milk expression and socio-demographic background. Infant characteristics were obtained from each infant's medical chart. 3.6 LABORATORY METHODS 3.6.1 Chemicals Chemicals used for lipid analysis were reagent grade and purchased from Fisher Scientific, Fisher Chemicals, Fair Lawn, New Jersey, USA and Sigma-Aldrich Chemical Company, Oakville, Ontario, Canada. Authentic lipid standards (C13:0 and C17:0) for direct methylation and gas liquid chromatography were purchased from Sigma-Aldrich Chemical Company, Oakville, Ontario, Canada. 3.6.2 Equipment Expressed breast milk samples were centrifuged using a low speed centrifuge, Sorvall® T6000B Centrifuge, with swinging basket rotor from Dupont (<6000 rpm) (Newton, CT, USA). Fatty acid methyl esters were separated and quantified by gas liquid chromatography (GLC), using a 3400 gas liquid chromatograph (Varian Canada Inc., Mississauga, Ontario), equipped with flame ionization detection and an IBM computer system using Varian "STAR" software. Creamatocrit measurements were obtained using a microhematocrit centrifuge, Haemofuge ®, American Scientific 37 Design and Methods Products, Heraeus-Christ GMbH, Germany. Creamatocrit values were determined using a CRIPTOCAPS ™ Micro-haematocrit Capilliary Tube Reader from Oxford Labware, Sherwood Medical, St. Louis, MO, USA (distributed by VWR CanLab, Mississauga, Ontario, Canada). 3.6.3 Lipid Analysis Three methods of lipid analysis were used to determine fat content of each mother's breast milk. Milk samples were analyzed by a standard solvent extraction method for gravimetric quantification, a simple micro-method that has clinical and practical applications, and by gas liquid chromatography for total fatty acids. 3.6.3.1 Gravimetric Analyses Human milk fat was extracted by a modified Folch method (Folch 1957), using the solvents chloroform/methanol/saline (8/4/3 v/v/v). Total fat content was quantified gravimetrically and is reported as grams fat per 100 mL milk. 3.6.3.2 Creamatocrit The Creamatocrit method, as described by Lucas et al. (1987), was used to estimate the fat content of human milk. A small amount of milk was drawn up into a hematocrit tube and then centrifuged to separate the milk sample into a cream layer and aqueous layer. The cream layer was measured and reported as a percentage of the total length of the milk column (creamatocrit value) using a creamatocrit reader. A calibration equation, in which total fat content of milk was: Fat (g/L) = (creamatocrit value - 0.59) - 0.146, developed by Lucas et al. (1987) was used to calculate the total fat content (g fat /100mL) from the creamatocrit value. During the method development phase of this study, initial tests were conducted using the Creamatocrit method to investigate the effects of freezing milk on the creamatocrit measurement. A sample of fresh and frozen expressed breast milk obtained from 1 mother was analyzed by the Creamatocrit method. The milk sample was frozen at -80°C for 2 months. This analysis revealed no significant difference (P>0.05) in mean creamatocrit value when comparing fresh and frozen milk samples (Table 3.1). Hence, frozen milk samples were used primarily in the study. 38 Design and Methods Table 3.1 Effect of Freezing Human Milk on the Creamatocrit Measurement1 Fresh Milk Frozen Milk Creamatocrit value (%)2 6.80% ± 0.26 6.85% ± 0.26 (mean ± SD) Coefficient of variation 3.8% 3.5% (CV %) 1 A fresh and frozen expressed breast friiik'sample obtained from one mother was analyzed using the Creamatocrit Method; results are shown for n = 10 measurements.; 2 Results were not significantly different (P>0.05) using the paired t=test. 3.6.3.3 Direct Methylation and Gas Liquid Chromatography (Fatty Acid Analysis) The fatty acid composition and total fatty acid content of breast milk were determined by a direct methylation procedure (Lepage 1988) and gas liquid chromatography. A 50 uL aliquot of EBM was used and C13:0 : C17:0 (1:1) added as internal standards. The sample was then methylated with 2 mL benzene/methanol (1/4 v/v) using the modified direct transesterification method of Lepage and Roy (1988). Acetyl-chloride (200uL) was added while vortexing, then the tubes were capped tightly and heated at 100°C for 60 minutes on a heating block. The tubes were vortexed every 15 minutes. The samples were neutralized by adding 5 mL of 6% K 2 C 0 3 , recapped and vortexed. Methyl esters were recovered with 6 mL pentane. The pentane layer was separated by centrifugation and removed, then dried under nitrogen gas. Samples were stored at -20°C until G L C analysis of fatty acid content. Separation of the fatty acid methyl esters was achieved using a gas liquid chromatograph equipped with a 100 m x 0.25 mm ID (internal diameter), 0.2 um film nonbonded, fused silica capilliary SP2560 columns (Supelco, Belefonte, PA). Helium was used as the carrier gas at a column flow of 1 mL/min and inlet pressure of 51 pounds per square inch. The inlet splitter was set at 10 to 1. Samples were injected at 100° C with the oven temperature programmed to remain at 100° C for 2 minutes, then increased to 190° C at 5° C/min, held for 20 min, then raised to 225 ° C at 10° C/min and held for 2 min. The column was then heated to 245° C at 25° C/min and held for 22 min before subsequent analyses. 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. Area under the curve for each fatty acid and retention times were recorded by a G L C data system (Varian 3400 with an 1MB computer system using Varian "STAR" software, Mississauga, Ontario). Total fatty acid 39 Design and Methods content was determined by the sum of all identified fatty acids (cis, trans and conjugated fatty acids). The medium chain fatty acid content was calculated as the sum of C8:0 -C12:0 fatty acids. The relative percentage of medium chain fatty acids, to total fatty acids was then calculated. The total fatty acid content of the milk was converted to total trigylceride and total fat by applying conversion factors. The total fatty acid content of milk (g/1 OOmL) was converted to trigylceride (TG) by applying a conversion factor of 1.05 (Jensen 1999). The total fatty acid content of milk (g/1 OOmL) was converted to total fat by applying a conversion factor of 1.063. These conversion factors assume that 98-99% of the total fat in milk is in the form of triglycerides and 1.3% of the remaining fat (lipid) is primarily present as phospholipids and cholesterol (Bitman 1983). Refer to Appendix 7 for more information about the conversion of fatty acids to total trigylceride and total fat content of milk. 3.6.4 Intra- and Inter- assay Variability Measurements for Lipid Analysis The intra and inter-assay variability measurements, coefficient of variation, for the modified Folch method and Creamatocrit method were determined by analyzing whole cows milk in 5 replicates over 3 separate days. The coefficients of variation were below 6% for all analyses, suggesting consistency in the lipid analysis methodology (Table 3.2). Table 3.2 Intra and Inter- Assay Variability Tests for the modified Folch Method of total lipid extraction and the Creamatocrit Method Intra-assay variability1 Coefficients of Variation (%) Inter-assay variability2 Coefficients of Variation (%) modified Folch method (frozen samples) 3.0 3.9 Creamatocrit method (fresh samples) 5.8 4.9 Creamatocrit method (frozen samples) 5.7 4.9 The intra-assay variability was determined in 5 samples of milk tested on the same day. 2The inter-assay variability was determined from the average values of 5 separate tests (each test consisted of 5 samples of milk) conducted over 3 days. 40 Design and Methods 3.7 DATA ANALYSIS Data were analyzed using the Statistical Package for the Social Sciences (SPSS Inc. Version 9.0 for Windows, Chicago, Illinois). Descriptive statistics were used to describe participant and infant characteristics, breast milk expression practices, fat content of milk, volume of milk expressed, practicality score, ranking of fractionating methods, and expressed breast milk practices survey data. Inferential statistics were used to test for significant differences between data sets. Results with a probability of less than 0.05 were considered statistically significant. All data were tested for normality using the Kolomogorov-Smirnov Test and normality plots (Q-Q plots) (Ntoumanis 2001). Data that did not satisfy the basic assumptions underlying the statistical test were modified by log transformation before statistical evaluation (these data included the foremilk volume, hindmilk volume, and percent hindmilk volume for all 3 fractionating methods) or had a correction factor applied (these data were the hindmilk volume) (Field 2000, Ntoumanis 2001, Zolman 1993). The following statistical tests were performed: 1) The paired-samples t-test was used to compare the means of the fat content of foremilk and hindmilk. 2) Repeated measures analysis of variance (ANOVA) was used to test for significant differences among the 3 fractionating methods (independent variable) for the fat content of milk, the volume of milk expressed and the practicality score of the methods (dependent variables). The repeated measures ANOVA test is a statistical procedure that involves taking measures on the same subjects who are exposed to 3 or more treatment/experimental conditions (Field 2000, Howell 1982, Munro 1993, Zolman 1993). The effect of order (random assignment) in performing the 3 fractionating methods was controlled for in the analysis. The Mauchly's Test of Sphericity was conducted to test the assumption of sphericity in the repeated measures ANOVA test. When the assumption was violated (hindmilk volume data) a correction factor, the Greenhouse and Geisser correction, was applied (Field 2000, Ntoumanis 2001). The Levene Test was conducted to test the assumption of homogeneous variances (Field 2000). When the assumption was violated (foremilk volume and % hindmilk volume data) the data were log transformed before statistical evaluation. 41 Design and Methods 3) The Friedman Test was used to test for significant differences among the 3 fractionating methods (independent variable) for data that were measured as ranks (dependent variable). 4) The Pearson correlation coefficient was used to examine potential relations between the fat content of milk (g/1 OOmL), as determined gravimetrically by the modified Folch method and the creamatocrit value, as determined by the Creamatocrit method. Potential relations between the volume of milk expressed and the fat content of milk were also examined. 5) Linear regression was used to evaluate the relation between the creamatocrit value, determined by the Creamatocrit method and the fat content of milk (g/1 OOmL), determined gravimetrically by the modified Folch method. A regression equation was developed to predict the estimated total fat content of milk (V) from the creamatocrit value (X). The formula for the linear equation was Y'=bX + a, where Y' was the predicted score, "a" was the point at which the line crosses the y-axis (intercept constant), "b" was the slope of the regression line (regression coefficient) and "X" was the known score or predictor variable (Munro 1993, Salkind 2000, Howell 1997). 42 Results CHAPTER 4 RESULTS The following chapter reports the findings from all 3 phases of this study. The first section of this chapter describes the results of the participant recruitment process and describes the characteristics of the participants and their infants. The second section summarizes the results of the practices of breast milk expression of mothers participating in each of the 3 study phases. In the next 3 sections, the results from each of the 3 phases are presented in sequential order in which the study was conducted, Phase I followed by Phase II and then Phase III. The following 2 sections of this chapter report the findings from the examination of the relationship between the 2 methods of fat analysis (gravimetric method and Creamatocrit Method) and the examination of the relationship between the fat content of milk and milk volume. The last section of this chapter summarizes the results of the study with respect to the study hypotheses. 4.1 PARTICIPANT RECRUITMENT AND CHARACTERISTICS 4.1.1 Phase I - Development of a protocol for fractionating expressed human milk into foremilk and hindmilk The recruitment period for Phase I was July 1999 to September 2000. A recruitment letter was left at the bedside of 55 infants who had mothers who were potential participants for this study. Of the 55 mothers identified, 23 (42%) agreed to participate in the study. Of the 23 participants recruited, 10 (43%) were unable to complete the study and were considered protocol failures. The reasons for protocol failure included: transfer of the infant to another hospital (n= 7), mother had an insufficient milk supply (n=1), infant was placed in strict MRSA (Methicillin-Resistant Staphylococcus aureus) isolation (milk samples could not be collected due to risk of contamination) (n=1), and mother lost interest in the study (n=1). Thirteen (57%) of the 23 participants recruited finished the study. Ten of these 13 successfully completed the entire study and are referred to as the completed group. Three of the 13 participants who finished the study were considered to be incomplete due to the following reasons: participants had missing data, missing milk samples, or their infant started breastfeeding (thereby affecting their pumping routine). The characteristics of the participants were obtained from the socio-demographic questionnaire and from their infant's medical chart (Table 4.1). Of the 23 mothers who participated in Phase I, 21 completed the socio-demographic questionnaire. Of the 21 43 Results mothers who completed the questionnaire, 2 participants did not answer all of the questions in the socio-demographic questionnaire. One mother did not provide information on her ethnic background and another mother did not provide information on the number of individuals dependent on the family income. Maternal age was obtained for all participants from their infant's medical chart. The mean age of the mothers who participated in Phase I was 28.6 ± 5.8 years (range 20-39 years, n=23). Seventeen of the 21 mothers had completed post-secondary education. Seventeen of the 21 mothers were married or living common-law. The study population consisted of 14 Caucasian mothers (70%), 1 East Indian mother (5%) and 4 First Nations mothers (20%). Seven of the 21 mothers (34%) had a gross annual family income of more than $50, 000 and 3 (14%) had a gross annual family income of less than $20,000. Eighteen of the 20 mothers had 3-5 persons living in their household (including their infant(s) in the hospital) who were dependent on the family income. Eleven of the 21 mothers had children, other than their infant(s) in the SCN, living in their family household. Nine of the 21 mothers lived in Vancouver or in the Lower Mainland/Fraser Valley area, while 12 mothers lived elsewhere in British Columbia. The mean number of days to complete the study for the mothers who successfully completed the baseline and 3 fractionating periods in Phase I was 18 ± 15 days (range 8 -59 days, n=10). The infant characteristics were obtained for Phase I from each infant's medical chart (Table 4.2). The mean gestational age of the infants at birth was 27 ± 2.5 weeks (range 23 - 32 weeks) and birth weight was 1041 ± 395 g (range 550 - 1900 g). The majority of the infants (n=20/23) were appropriate for gestational age (AGA) with a birth weight between the10-90%ile for age. For the 3 mothers who gave birth to twins, the average weight of 2 infants was used for analysis. At the time of the mothers' recruitment into the study, the infants had a mean corrected gestational age of 32 ± 2.5 weeks (range 28-37 weeks). The mean day of life of the infants at recruitment was 35 ± 12 days (range 1 8 - 5 9 days). The mean weight of the infants at recruitment was 1389 ± 403 g (range 780 - 2225 g). 44 Results Table 4.1 Characteristics of Participants in Phase I1 n %_ Maternal Age (y) (n=23) 2 0 - 2 5 9 39 2 6 - 3 0 5 22 31 - 3 5 7 30 >36 2 9 Mean ± SD (range) 28.6 ± 5.8 (20 - 39) Highest Level of Education Completed (n=21) Less than High School 1 5 High School 3 14 Community College/Technical School 11 52 Undergraduate Degree 4 19 Graduate Degree 1 5 Other (private college) 1 5 Marital Status (n=21) Single 3 14 Married / common-law 17 81 Separated 1 5 Ethnic Background /Race (n=20) East Indian 1 5 Caucasian 14 70 First Nations 4 20 Other 1 5 Gross Annual Family Income (n=21) < $20,000 3 14 > $20,000 - $50,000 11 52 > $50,000 7 34 Number of Individuals Dependent on Family Income2 (n=20) <2 1 5 3 - 5 18 90 >6 1 5 Children Living in Household 3 (n=21) Yes 11 52 No 10 48 City or Town of Residence 4 (n=21) Vancouver/Richmond 4 19 Lower Mainland / Fraser Valley Region 5 5 24 Communities Elsewhere in B C 6 12 57 Maternal age was obtained for all participants from their infant's medical chart. Of the 23 mothers who participated in Phase I, 21 completed the socio-demographic questionnaire. Of the 21 mothers who completed the questionnaire, 2 mothers did not answer all of the questions on the socio-demographic questionnaire. 2 Includes infant(s) in hospital 3 Does not include infant(s) in hospital 4 Exception: Winlaw is not designated a town or city, it is unincorporated. 5Abbotsford, Surrey, Port Moody, Chilliwack; 6 Naniamo, Duncan, Terrace, Prince George, Kamloops, Hazelton, Okanagon Falls, Winlaw, Creston 45 Results Table 4.2 Characteristics of Participants' Infants in Phase I1 BirthWeight (g) < 1000 1001 - 1500 1501 - 2000 Mean ± SD (range) Gestational Age at Birth (wks) <25 26-30 31-35 Mean ± SD (range) Size for Gestational Age at Birth Small (birth weight < 10%ile) Appropriate (birth weight >10 & < 90%ile) Large (birth weight > 90%ile) Weight at Recruitment (g) < 1000 1001 - 1500 1501 - 2 0 0 0 >2001 Mean ± SD (range) Corrected Gestational Age at Recruitment (wks) 26-30 31-35 >36 Mean ± SD (range) Day of Life at Recruitment (d) <21 2 2 - 3 5 3 6 - 4 9 5 0 - 6 3 Mean ± SD (range) n= 23 14 7 2 1041 ± 3 9 5 (550-1900) 7 13 3 27 ± 5.5 (23 - 32) 2 20 1 5 9 7 2 1389 ± 4 0 3 (780 - 2225) 7 15 1 32 ± 2 . 5 (28 - 37) 4 9 7 3 35 ± 12 (18-59) % 61 30 9 30 57 13 9 87 4 22 39 30 9 30 65 4 17 39 30 13 Twenty-three mothers participated in Phase I of the infant's medical chart. Characteristics of the participants' infant were obtained from each 46 Results 4.1.2 Phase II - Reproducibility of a protocol for fractionating expressed human milk into foremilk and hindmilk The recruitment period for Phase II was from January 2001 to July 2001. A recruitment letter was left at the bedside of 31 infants who had mothers who were potential participants in this study. Of the 31 mothers identified, 8 (26%) mothers agreed to participate in the study. All 8 mothers completed the 24-hour baseline collection period. Two of the 8 mothers were unable to complete the entire study and were considered protocol failures because their infants were transferred to another hospital. Six of the 8 mothers (75%) recruited for Phase II completed the study. Four of the 6 mothers were able to successfully complete 2 fractionating days and 2 mothers were able to successfully complete all 3 fractionating days. The data collected from the fractionating days were averaged for each mother, and then used in the analysis. The characteristics of the participants for Phase II were obtained from the socio-demographic questionnaire and from their infant's medical chart (Table 4.3). All the mothers (n=8) in Phase II completed the socio-demographic questionnaire. The mean age of the mothers in Phase II was 28.6 ± 5.7 years (range 20-38 years, n=8). Six (75%) of the mothers had completed post-secondary education. All the mothers were married or living common-law. The study population consisted of 6 Caucasian mothers (75%) and 2 East Indian mothers (25%). Three (37.5%) mothers had a gross annual income of >$50,000, 4 (50%) mothers had a gross income of $20,000-$50,000 and 1 (12.5%) mother had a gross annual income of <$20,000. All the mothers had 3-5 persons living in their household (including their infant(s) in the hospital) who were dependent on the family income. Three (38%) of the mothers had children, other that their infant(s) in the SCN, living in their family household. Three of the mothers (37.5%) lived in Vancouver/Lower Mainland and 2 (25%) of the mothers lived elsewhere in British Columbia. Three (37.5%) of the mothers lived in communities that were outside of British Columbia. The mean number of days to complete the study with at least baseline and 2 fractionating days for the mothers in Phase II was 11 ± 7 days (range 7-25 days). The infant characteristics for Phase II were obtained from each infant's medical chart (Table 4.4). The mean gestational age of the infants at birth was 28 ± 2.4 weeks (range 25-32 weeks) and birth weight was 909 ± 300 g (range 545-1505 g). Most of the infants (n=6/8) were appropriate for gestational age (AGA), with a birth weight between 47 Results the10-90%ile for age. One participant gave birth to twins and the average weight of the 2 infants was used for analysis. At the time of the mothers recruitment into the study, the mean corrected gestational age of the infants was 32 ± 2 weeks (range 30-35 weeks). The mean day of life of the infants at recruitment was 34 ± 18 days (range 17-66 days). The mean weight of the infants at recruitment was 1342 ± 417 g (range 700 -1970 g). 48 Results Table 4.3 Characteristics of Participants in Phase II1 n=8 o/ la Maternal Age (y) 2 0 - 2 5 2 6 - 3 0 31 - 3 5 >36 Mean ± SD (range) Highest Level of Education Completed High School Community College/Technical School Undergraduate Degree Marital Status Married / common-law 3 2 2 1 28.6 ± 5.7 (20 - 38) 2 2 4 37.5 25 25 12.5 25 25 50 100 Ethnic Background / Race East Indian Caucasian 2 6 25 75 Gross Annual Family Income < $20, 000 > $20,000 - $50,000 > $50,000 1 4 3 12.5 50 37.5 Number of Individuals Dependent on Family Income2 3 - 5 100 Children Living in Household Yes No 3 5 38 62 City or Town of Residence Vancouver Lower Mainland Region 4 Communities elsewhere in B C 5 Communities Outside B C 6 2 1 2 3 25 12.5 25 37.5 All the mothers who participated in Phase II completed the socio-demographic questionnaire. ' Includes infant(s) in hospital 3 Does not include infant(s) in hospital "Surrey; 5 Hazelton, Kelowna; 6Lethbridge,Whitehorse 4 9 Results Table 4.4 Characteristics of Participants' Infants in Phase II1 n=8 /o Birth Weight (g) < 1000 1001 - 1 5 0 0 1501-2000 Mean ± SD (range) Gestational Age at Birth (wks) <25 26-30 31-35 Mean ± SD (range) Size for Gestational Age at Birth Small (birth weight < 10%ile) Appropriate (birth weight >10 & < 90%ile) Weight at Recruitment (g) < 1000 1001 - 1500 1501 - 2000 Mean ± SD (range) Corrected Gestational Age at Recruitment (wks) 26-30 31-35 Mean ± SD (range) Day of Life at Recruitment (d) <21 2 2 - 3 5 3 6 - 4 9 5 0 - 6 3 >63 Mean ± SD (range) 5 2 1 909 ± 300 (545-1505) 1 6 1 28 ± 2.4 (25 - 32) 2 6 1 4 3 13421417 (700-1970) 2 6 32 ± 2 . 0 (30 - 35) 1 5 1 1 34 ± 18 (17-66) 62.5 25 12.5 12.5 75 12.5 25 75 12.5 50 37.5 25 75 12.5 62.5 12.5 12.5 Eight mothers participated in Phase II. Characteristics of the participants' infant were obtained from each of the infant's medical chart. 50 Results 4.1.3 Phase III - Determination of expressed breast milk practices and volumes The recruitment period for Phase III was from May 2001 to December 2001. A recruitment letter was left at the bedside of 33 infants who had mothers who were potential participants. Of the 33 mothers identified, 19 (58%) mothers agreed to participate in the study. Six of the 19 participants were unable to complete the study and were considered protocol failures. The reasons for protocol failures included: transfer of the infant to another hospital (n= 4), or the infant started to breastfeed, therefore, the mother did not express her milk (n=2). Thirteen of the 19 mothers completed the 3-day milk expression diary. Twelve of the 13 mothers successfully completed a 3-day milk expression record. For 1 of the 13 mothers, a 2-day milk expression record was used because the infant started breastfeeding on the last (third) recording day, thereby affecting the mother's pumping routine and the volume expressed. The data collected from the milk expression records were averaged for each mother, and then used in the analysis. The characteristics of the participants for Phase III were obtained from the socio-demographic questionnaire and from their infant's medical chart (Table 4.5). Of the 19 mothers who completed Phase III, 15 mothers completed the socio-demographic questionnaire. One of the 15 mothers did not provide information on her annual family income and another mother did not provide information on her ethnic background. Maternal age was obtained for all the participants from their infant's medical chart. The mean age of the mothers in Phase III was 29.8 ± 6.0 years (range 20-42 years, n=19). All the mothers who completed the questionnaire (n=15/19) had completed post-secondary education. Most of the mothers (93%) were married or living common-law and only one mother was single. The study population consisted of 9 Caucasian mothers (64%), 1 Chinese mother (7%), 1 First Nations mother (7%), and 1 Latin American mother (7%). Eight (57%) of the 14 mothers had a gross annual income of > $50,000, 5 mothers (36%) had a gross income of $20,000-50,000 and 1 (7%) mother had a gross income of <$20,000. Most of the mothers (80%) had 3-5 persons living in their household (including their infants in the hospital) who were dependent on the family income. Six (40%) of the 15 mothers had children, other than their infant(s) in the SCN, living in their family household. Nine of the 15 mothers lived in Vancouver or in the Lower Mainland/Fraser Valley area, while 7 mothers lived elsewhere in British 51 Results Columbia. The mean number of days for the mothers to complete the study with at least 2 of the 3 reporting days was 14 ± 14 days (range 5-61 days, n=13). The infant characteristics were obtained for Phase III from each infant's medical chart and from the participant's study booklet (Table 4.6). The mean gestational age of the infants at birth was 29 ± 2 weeks (range 26-32 weeks) and the mean birth weight was 1183 ± 396 g (range 590 - 2460 g). The majority of the infants (n=16/19) were appropriate for age (AGA), with a birth weight between the 10-90%ile for age. Five participants delivered twins and the average weight of the 2 infants was used for analysis. At the time of the mother's recruitment into the study, the mean corrected gestational age of the infants was 34 ± 2.1 weeks (range 29-38 weeks). The mean day of life of the infants at recruitment was 34 ± 15 days (range 14-67 days). The mean weight of the infants at recruitment was 1707 ± 447g (range 590 - 2355 g). 52 Resul ts Table 4.5 Participant Characteristics in Phase III1 n % Maternal Age (y) (n= 19) <20 1 5 2 0 - 2 5 4 21 2 6 - 3 0 7 37 31 -35 2 11 > 36 5 26 Mean ± SD (range) 29.8 ± 6.0 (20-42) Highest Level of Education Completed (n=15) Community College/Technical School 11 73 Undergraduate Degree 3 20 Graduate Degree 1 7 Marital Status (n=15) Single 1 7 Married / common-law 14 93 Ethnic Background/ Race (n= 14) Chinese Caucasian First Nations Latin American Other 1 7 9 64 1 7 1 7 2 14 Gross Annual Family Income (n=14) < $20,000 1 7 > $20,000 - $50,000 5 36 > $50,000 8 57 Number of Individuals Dependent on Family Income2 (n=15) <2 2 13 3 - 5 12 80 >6 1 7 Children Living in Household 3 (n=15) Yes 6 40 No 9 60 City or Town of Residence 4 (n=15) Vancouver / Richmond 5 33 Lower Mainland / Fraser Valley Region 5 3 20 Communities elsewhere in B C 7 47 Of the 19 mothers who participated in Phase III, 15 completed the socio-demographic questionnaire. Maternal age was obtained for all participants from their infant's medical chart. Two mothers did not answer all of the questions in the socio-demographic questionnaire, inc lud ing infant(s) in hospital 3 Does not include infant(s) in hospital 4 Exception: Whistler is designated a resort municipality. 5Abbotsford, Burnaby; 6 Ladysmith, Altin, Prince George, Kelowna, Vernon, Revelstoke, Whistler 53 Resul ts Table 4 .6 Characteristics of Participants' Infants in Phase III1 BirthWeight (g) < 1000 1001 -1500 1501 -2000 >2001 Mean ± SD (range) Gestational Age at Birth (wks) 26-30 31-35 Mean ± SD (range) n=19 % 7 37 10 53 1 5 1 5 1183 ±396 (590 - 2460) 16 84 3 16 29 ±2.1 (26 - 32) Size for Gestational Age at Birth Small (birth weight < 10%ile) 2 11 Appropriate (birth weight >10 & < 90%ile) 16 84 Large (birth weight > 90%ile) 1 5 Weight at Recruitment (g) <1000 1 5 1001 -1500 4 21 1501 - 2000 9 48 >2001 5 26 Mean ± SD (range) 1707 ± 447 (590 - 2355) Corrected Gestational Age at Recruitment (wks) 26-30 2 11 31-35 16 84 >36 1 5 Mean ± SD (range) 34 ±2.1 (29 - 38) Day of Life at Recruitment (d) <21 6 32 22-35 4 21 3 6 - 4 9 5 26 5 0 - 6 3 2 11 >63 1 5 Mean ± SD (range) 34 ±15 (14-67) Nineteen mothers participated in Phase III. Characteristics of the participants' infant were obtained from each of the infant's medical chart and from the participant's study booklet. 54 Resul ts 4.2 PRACTICES OF BREAST MILK EXPRESSION 4.2.1 Practices of Breast Milk Expression During the Baseline Period in Phase I The results of the pract ices of breast milk express ion during the basel ine period in P h a s e I are summar ized in Table 4.7. Twenty-one mothers completed the 24-hour basel ine period. Two of the 21 mothers did not report information on the duration of pumping and 1 mother did not report information on the method of pumping. The mean frequency of pumping (number of pumping sess ions per day) in the basel ine period was 6 ± 1 (range 4-8 sess ions per day). The mean duration of pumping (total minutes pumping per day) for the basel ine period was 156 + 45 minutes (range 80-240 minutes). Thirteen of the 20 mothers used the double pumping method for col lection of exp ressed milk. In a double pumping method, mothers pump both breasts at the s a m e time during a single pumping sess ion (simultaneous pumping). In a single pumping method, mothers pump 1 breast at a time, fol lowed by the second breast during a single pumping sess ion (sequential pumping). For mothers who double pumped, the pumping time was doubled to obtain the total minutes pumped during that sess ion . Mothers who double pumped had a mean pumping time of 15 minutes per sess ion (total pumping time was 30 minutes) and mothers who single pumped had a mean pumping time of 23 minutes per sess ion . Al l the mothers used an electric pump and pumped both breasts at each sess ion . Table 4.7 Practices of Breast Milk Expression During the Baseline Period in Phase I1 Number of Mothers Mean ± SD Range Frequency of pumping (number of pumping sessions per day) 21 6 ± 1 4 - 8 Duration of pumping2 (total minutes of pumping per day) 19 156 ± 4 5 80 - 240 Method of pumping3 Single pumping Double pumping Both 20 6 13 1 Twenty-one mothers completed the baseline period. Information on pumping duration was not reported by 2 mothers and method of pumping was not reported by 1 mother. 2 Duration of a pumping session was calculated as the sum of pumping time for both breasts at a pumping session. For mothers who double pumped, the pumping time was doubled to obtain the total minutes pumped. 55 Results 4.2.2 Practices of Breast Milk Expression During the Baseline Period in Phase II The results of the practices of breast milk expression during the baseline period in Phase II are summarized in Table 4.8. All of the mothers (n=8) completed the baseline period. One of the 8 mothers did not report information on the duration of pumping or the method of pumping. The mean frequency of pumping in the baseline period was 5 ± 1 (range 4-7 sessions per day). The mean duration of pumping in the baseline period was 120 ± 42 minutes per day (range 42 -180 minutes). For mothers who double pumped, the pumping time was doubled to obtain the total minutes pumped during that session. Four mothers used the double pumping method; while 1 mother used the single pumping method and 2 mothers used a combination of the 2 methods. Mothers who double pumped had a mean pumping time of 12 minutes per session (total pumping time was 24 minutes), and the mother who single pumped had a mean pumping time of 20 minutes per session. All the mothers used an electric pump for expressing milk, however, 1 mother also used a hand pump. All the mothers pumped both breasts at each pumping session. Table 4.8 Practices of Breast Milk Expression During the Baseline Period in Phase II1 Number of Mean + SD Range Mothers Frequency of pumping 8 5 + 1 4-7 (number of pumping sess ions per day) Duration of pumping2 7 120 ± 4 2 42-180 (total minutes of pumping per day) Method of pumping 7 Single pumping 1 Double pumping 4 Both 2 Eight mothers completed the baseline period. Information on pumping duration and method of pumping was not reported by 1 mother. 2 Duration of a pumping session was calculated as the sum of pumping time for both breasts at a pumping session. For mothers who double pumped, the pumping time was doubled to obtain the total minutes pumped. 56 Results 4.2.3 Practices of Breast Milk Expression in Phase III The results of the practices of breast milk expression of mothers in Phase III are summarized in Table 4.9. Thirteen mothers completed a milk expression diary in Phase III. Twelve of the 13 mothers successfully completed a 3-day milk expression record and 1 mother successfully completed a 2-day milk expression record. The data collected from the milk expression records were averaged and used to describe the practices of breast milk expression for each mother. The mean frequency of pumping per day was 6 ± 1 (range 4-7 sessions per day). The mean duration of pumping was 135 ± 58 minutes per day (range 61-257 minutes). For mothers who double pumped, the pumping time was doubled to obtain the total minutes pumped during that session. More than half of the mothers used the double pumping method, while 1 mother used the single pumping method and 5 mothers used a combination of the 2 methods. Mothers who double pumped had a mean pumping time of 11 minutes per session (total pumping time was 22 minutes) and the 1 mother who single pumped had a mean pumping time of 20 minutes per session. All the mothers used an electric pump, however, 1 mother also used a hand pump. All the mothers pumped both breasts at each pumping sessions. Table 4.9 Practices of Breast Milk Expression in Phase III1 Mean ± SD Range n=13 Frequency of pumping 6 ± 1 4-7 (number of pumping sessions per day) Duration of pumping2 135 ± 5 8 61-257 (total minutes of pumping per day) Method of pumping Single pumping 1 Double pumping 7 Both 5 Thirteen mothers completed the milk expression diary in Phase III. Twelve mothers completed a 3-day milk expression record and one mother completed a 2-day milk expression record. The data collected from the milk expression record was averaged and used to describe the expressed breast milk practices for each mother. 2 Duration of a pumping session was calculated as the sum of pumping time for both breasts at a pumping session. For mothers who double pumped, the pumping time was doubled to obtain the total minutes pumped. 57 Results 4.3 PHASE I - DEVELOPMENT OF A PROTOCOL FOR FRACTIONATING EXPRESSED HUMAN MILK INTO FOREMILK AND HINDMILK 4.3.1 Fat Content of Preterm Milk The fat content of breast milk expressed by the 19 mothers of premature infants (preterm milk) who collected during the baseline period in Phase I is shown in Table 4.10. The mean fat content of mothers' milk (n=19) was 3.95 ± 0.83 g/1 OOmL, as determined gravimetrically by the modified Folch method (Folch 1965). The range of fat content was 2.33 - 5.40 g /1 OOmL. The mean fat content of milk for the 10 mothers who completed Phase I successfully was 3.91 ± 0.74 g/100 mL, with a range of 2.33 to 5.09 g/1 OOmL (Table 4.10). The mean creamatocrit value for the baseline milk samples (n=19) was 7.19 ± 1.33%, equivalent to 4.52 ± 0.91 g fat/1 OOmL, using the equation determined by Lucas et al (1978). For the subgroup of mothers who completed Phase I of the study (n=10), the mean creamatocrit value was 7.40 ± 1.52 %, corresponding to 4.66 ± 1.0 g fat/1 OOmL. The mean total fatty acid content of the baseline milk samples (n=19) was 3.85 ± 0.77 g fat/1 OOmL. For mothers who completed Phase I successfully, the mean total fatty acid content of milk was 3.80 ± 0.52 g fat/1 OOmL, n=10. The mean triglyceride (TG) content of all the baseline milk samples was 4.05 ± 0.81 g fat/1 OOmL, n=19 (1.05 x total fatty acid content of milk), and 3.99 ± 0.54 g fat/1 OOmL for the 10 women who completed Phase I of the study. The mean total fat content of milk determined by analysis of total fatty acids using G L C (total fatty acids x 1.063) was 4.10 ± 0.82 g/1 OOmL for the total group (n=19), and 4.04 ± 0.55 g/1 OOmL for the completed group (n=10). 58 Results Table 4.10 Fat Content of Preterm Milk in Phase I1 Fat Analysis Method Total Group (n=19) Completed Group (n=10) Mean ± SD Range Mean ± SD Range Gravimetric Method 2 (g/1 OOmL) 3.95 ± 0.83 2.33 - 5.40 3.91 ± 0.74 2.33-5.09 Creamatocrit Method 3 (g/1 OOmL) 4.52 ±0.91 2 . 5 6 - 5 . 7 6 4.66 ± 1.00 2.56-5.76 Fatty Acid Method 4 4.10 ±0.82 2.80- 5.79 4.04 ± 0.55 2.90 - 4.64 (g/1 OOmL) Results are shown for the baseline period in Phase I; 19 mothers in the total group and 10 mothers in the completed group. 2Fat extracted by modified Folch method (1956) and quantified gravimetrically. 3 The Creamatocrit value was 7.19 + 1.33% (mean ± SD), range 4.33-9.00%. The fat content was calculated from the creamatocrit value using the Lucas Equation (1978): [fat (g/L)=(creamatocrit value - 0.59)/0.146] Tatty acid content was determined by gas liquid chromatography and then converted to triglyceride (TG) and total fat by applying the conversion factors 1.05 and 1.063, respectively (assumes 98-99% of the total milk fat is in the form of TG). The total fatty acid content was 3.85 ± 0.77 g/1 OOmL (mean ± SD), range 2.63 - 5.44g/100mL. The TG was 4.05 ± 0.81 g/1 OOmL (mean ± SD), range 2.76-5.71 g/1 OOmL. 4.3.2 Fat content of Foremilk and Hindmilk Ten mothers successfully completed Phase I by separating their expressed breast milk into foremilk and hindmilk using the three fractionating methods. The fat content of foremilk and hindmilk determined gravimetrically for each method is summarized in Table 4.11. The fat content of foremilk and hindmilk separated using the Visual Method was 2.51 ± 0.53 g/1 OOmL and 4.68 ± 0.68 g/1 OOmL, respectively. The fat content of foremilk and hindmilk separated by the Time Method was 2.70 ± 0.72 g/1 OOmL and 4.96 ± 0.82 g/1 OOmL, respectively. Milk separated using the Volume Method produced a foremilk with a fat content of 2.99 ± 0.78 g/1 OOmL and a hindmilk with a content of 5.18 ± 0.96 g/1 OOmL. There was a significantly higher (P<0.001) fat content in hindmilk compared with foremilk for each of the 3 fractionating methods. 59 Results Table 4.11 Fat Content (g/1 OOmL) of Foremilk and Hindmilk Separated Using Three Fractionating Methods 1 2 Visual method4 Time method5 Volume method6 P value3 Foremilk (g fat/1 OOmL) 2.51 ± 0.53a (1.71-3.08) 2.70 ± 0.72a (1.57-3.65) 2.99 ± 0.78a (1.41-4.09) 0.226 Hindmilk (g fat/1 OOmL) 4.68 ± 0.60b (3.59-5.60) 4.96 ± 0.82b (3.31-6.05) 5 .18±0.96 b (3.08-6.75) 0.354 Difference in fat7 (g fat/1 OOmL) 2.17 ±0.62 (1.40-3.25) 2.26 ± 0.52 (1.65-3.16) 2.19 ±0.42 (1.64-2.85) 0.836 Increase Fold 8 1.92 ±0.37 (1.45-2.38) 1.91 ±0.36 (1.50-2.66) 1.78 ±0.25 (1.51 -2.18) 0.239 Mean ± SD (range) 2 Values are reported for mothers who successfully completed Phase I (Completed Group), n=10. 3There were no differences among the three methods of milk fractionation in the fat content of foremilk, hindmilk or in the difference in fat content between foremilk and hindmilk using Repeated Measures ANOVA. Visual Method: point of separation of foremilk and hindmilk is based on the subjective observation of when a change in color and /or consistency of milk occurs. 5 Time Method: point of separation of foremilk and hindmilk is based on 2-3 minutes from letdown or the start of a pumping session. 6 Volume Method: point of separation of foremilk and hindmilk is based on a proportion (%value) of mother's milk volume expressed at a pumping session. 7The difference in fat refers to the difference in fat content from foremilk to hindmilk. 8 The increase fold was calculated by dividing the hindmilk fat content by foremilk fat content. a bValues within a column with a different superscript for foremilk and hindmilk are different (P<0.001); paired t-test. There was no significant difference in the fat content of foremilk or hindmilk, or in the relative increase in fat in hindmilk compared to foremilk separated by the 3 fractionating methods (Table 4.11). The fat content increased by a factor of 1.8 to 1.9 from foremilk to hindmilk for all 3 methods. 60 Results 4.3.3 Volume of Expressed Preterm Milk 4.3.3.1 Volume of Expressed Preterm Milk During the Baseline Period The volume of breast milk expressed by mothers during the 24-hour baseline period in Phase I is summarized in Table 4.12. The mean volume of milk expressed by mothers was 783 ± 298 mL per day (range 293-1435 mL, n=20). Six mothers (5 mothers who delivered singletons and 1 mother who delivered twins) expressed more than 1000 mL milk per day. All mothers produced more than their infant's enteral intake at the time of the baseline collection. The percentage of milk volume expressed by mothers relative to their infant's intake ranged from 122% to 664% (n=20). Similar results were obtained for the group of mothers who successfully completed Phase I. The mean volume of milk expressed was 786 ± 276 mL per day (range 389 -1262 mL, n=10). The percentage of milk volume expressed by mothers relative to their infant's intake ranged from 126% to 633% (n=10). Table 4.12 Volume of Expressed Preterm Milk During the Baseline Period n Milk volume Range (mL/day) Mean ± SD Total Group1 20 783 ± 298 293-1435 Completed Group2 10 786 ± 276 389-1262 Total Group refers to the 20 mothers who completed the baseline period in Phase I. 2Completed Group refers to the 10 mothers who successfully completed Phase I. 4.3.3.2 Foremilk and Hindmilk Volumes Ten mothers successfully completed Phase I by separating their expressed breast milk into foremilk and hindmilk using all 3 fractionating methods. The volume of foremilk and hindmilk expressed using the three methods are shown in Table 4.13. The mean volumes of foremilk ranged from 352 to 447 mL and were not significantly different among the 3 fractionating methods. The mean volumes of hindmilk ranged from 320 to 446 mL and were not significantly different among the 3 fractionating methods. The mean volumes of hindmilk produced as a percentage of the total volume of milk expressed by mothers (hindmilk volume divided by the total volume x 100%) 61 Results ranged from 46 - 56% for the methods, with the Volume Method having the lowest hindmilk value at 46%. The number of mothers whose hindmilk volume was unable to meet their infant's enteral intake is illustrated for each of the 3 fractionating methods in Figure 4.1. The Volume Method had the highest number of mothers (n=4) who were unable to express enough hindmilk to meet their infant's enteral intake. Two mothers were unable to meet their infant's intake when fractionating milk using the Time Method. One mother was unable to meet her infant's intake when fractionating milk using the Visual Method. The percent of hindmilk volume expressed by mothers relative to their infant's intake [(hindmilk volume -^infant's intake) x 100%] ranged from 71% to 257% for the Visual Method, 47% to 494% for the Time Method and 19% to 162% for the Volume Method. Table 4.13 Comparison of Volumes of Foremilk and Hindmilk Obtained Using Three Methods of Fractionating Expressed Milk 1' 2 Visual method Time method Volume method P value 3 Foremilk (mL/ day) 3 5 9 ± 1 8 3 (183-631) 352 ±129 (199-600) 447 ± 295 (100-978) 0.696 Hindmilk (mL/day) 430 ± 205 (205-720) 446 ±217 (136-830) 320 ±167 (33-625) 0.795 Proportion of total milk as Hindmilk 4 (%) 56 ± 8 (42-69) 54 ± 10 (55-65) 46 ± 2 3 (4-83) 0.305 Proportion of hindmilk volume relative to infant intake 5 (%) 147 ± 5 9 (71-257) 170 ±131 (47-494) 102 ± 3 7 (19-162) 0.369 Mean ± SD (range) 2 Values are shown for mothers who successfully completed Phase I, n=10. 3 N o significant difference among methods using Repeated Measures A N O V A . "Hindmilk volume is reported as a percentage of total volume expressed [(hindmilk volume total volume) x 100%]. 5Hindmilk volume is reported as a percentage of infant intake [(hindmilk volume infant intake) x 100%]. 62 Results i) Visual method £ 300 CO 0 Study Participants ii) Time method iii) Volume method Study Participant Figure 4.1 Ability of Volume of Mothers' Expressed Hindmilk to Meet Infants' Intake1,2 ' E a c h b a r s r e p r e s e n t 1 m o t h e r . T h e he igh t of the ba r r e p r e s e n t s e a c h m o t h e r ' s v o l u m e of h i ndm i l k a s a p e r c e n t a g e of h e r in fant 's i n take [ (h indmi lk v o l u m e +infant 's in take) x 1 0 0 % ] . 2 R e s u l t s a r e s h o w n for m o t h e r s w h o s u c c e s s f u l l y c o m p l e t e d P h a s e I, n=10 63 Results 4.3.4 Evaluation Summary of Fractionating Methods 4.3.4.1 Practicality Scores of the Fractionating Methods Mothers who successfully completed Phase I evaluated the 3 fractionating methods, and from these evaluations a preferred method was determined. The mothers rated each method on a scale of 1 to 5 for practicality using 3 criteria: convenience, clear directions and time to complete the method. A practicality score (maximum score of 15) was determined for each method. The Time Method had the highest mean practicality score of 12.7 ± 1.2, the Visual Method had the second highest mean score at 11.8 ± 2.0 and the Volume Method had the lowest mean score of 10.9 ± 1.7. Comparison of the 3 scores, using repeated measures ANOVA showed no significant difference among methods (Table 4.14). Table 4.14 Comparison of Practicality Score for Three Fractionating Methods1 Practicality Criteria2 Visual method Time method Volume method Convenience3 3.1 ±1.2 3.8 ± 1.0 3.0 ±1.2 Clear Directions4 4.7 ±0.7 4.9 ±0.3 3.9 ±0.9 Time to complete5 4.1 ±0.8 4.0 ± 0.7 4.0 ±0.8 Practicality Score6'7 11.8 ±2.0 12.7 ±1.2 10.9± 1.7 Results are shown for mothers who successfully completed Phase I, mean score ± S D , n=10 2 Mothers rated each practicality criteria on a scale of 1 to 5; results shown are mean values. 3 Rating scale: a value of 1 corresponds to inconvenient and a value of 5 corresponds to very convenient. 4 Rating scale: a value of 1 corresponds to unclear directions and a value of 5 corresponds to very clear directions. 5 Rating scale: a value of 1 corresponds to significantly more time than their usual expression time and a value of 5 corresponds to no difference in time compared to usual expression time. 6 Practicality score was determined for each method as the sum of the 3 criteria scores; maximum score of 15. 7 No significant difference in practicality score among methods (P=0.096) using Repeated Measures ANOVA. 4.3.4.2 Preference Rankings of the Fractionating Methods The mothers ranked the 3 fractionating methods from the most preferred method to the least preferred method on a scale of 1 to 3. A value of 1 represented the most preferred method and a value of 3 represented the least preferred method. The Time Method was ranked as the most preferred by more mothers (6 out of 10 mothers) than the other 2 fractionating methods (Figure 4.2). The Visual and Volume Method had 64 Results similar first rankings (2 out of 10 mothers). Five mothers ranked the Volume Method as the least preferred method, 4 out 10 mothers ranked the Visual Method as the least preferred method, and only 1 mother ranked the Time Method as the least preferred method. The mean ranking value was 2.2, 1.5 and 2.3 for the Visual Method, Time Method and Volume Method, respectively. There was no significant difference (P=0.150) when comparing the mean ranking values among the methods, as determined by the Friedman Test. </> o 7 6 £ 5 in if4 C II o >> o c Q) 3 CT 0) 3 ^ 0 Preference Ranking Figure 4.2 Frequency of Preference Ranking for Three Fractionating Methods 1,2,3 Results are shown for mothers who successfully completed Phase I (Completed Group), n=10. 2 Bars represent the number of mothers ranking each method on a scale of 1 to 3, where 1 represents the most preferred method and 3 represents the least preferred method. 3 Visual Method | ; Time Method Q ; Volume Method \Z2 65 Results i) Visual method ii) Time method iii) Volume method Figure 4.3 Comparison of Willingness to Perform Fractionating Method1'2 1Results are for mothers who successfully completed Phase I, n=10 2Values shown are the number of mothers; DK=Don't know 66 Results 4.3.4.3 Mothers' Willingness to Perform the Fractionating Methods The willingness of the mothers to perform the fractionating method for an extended time period is illustrated in Figure 4.3. Six of the 10 mothers were willing to perform the Visual Method for an extended period of time. Eight of the 10 mothers were willing to perform the Time Method. Seven of the 10 mothers were willing to perform the Volume Method. Three of the 10 mothers were not willing to perform the Visual Method for an extended period of time. One of the 10 mothers was not willing to perform the Volume Method. The Time Method had the highest number of mothers willing to perform the method, while the Visual Method had the highest number of mothers not willing to perform the method. 4.3.4.4 Mothers' Comments on the Fractionating Methods Mothers provided comments on the practicality of the fractionating methods and general feedback on the collection procedures. The following are selected comments that reflect the mothers' preferences for separating milk into foremilk and hindmilk. Visual Method (011) "Would be very convenient if change of color were obvious." (013) "Best forme. Most flexible method. (No calculator or watch needed)." (014) "Couldn't see while flowing, so just picked a point." (023) "As already mentioned, this method may be difficult because transition from fore to hindmilk secretion may be gradual and not obviously noticeable." Time Method (007) "Time was very clear, easy compared to others." (014) "Like it best. ... Milk came within the 2-3 minutes therefore 1-2 minutes separation might have been better." (019) "Time, easier to do; less thinking." (022) " Time, because it is a general half way point and it is not as straining, because the other ones you need to always keep eye on bottle and it is bad for your neck." 67 Results Volume Method (001) "After the math was done for this method, overall this was the easiest for me, and I liked this one the best." (007) "It may be worthwhile taking an average of several baseline days recorded when getting the foreihindmilk ratio..." (011) "Need books, pens and paper, very inconvenient, (not very portable). Takes good direction/math skills." (022) "Probably the most logical for the study, but milk volume changes day to day therefore cannot be exact." 4.3.4.5 Summary for the Determination of an Acceptable Fractionating Method By Mothers Although there were no significant differences found when comparing the mean practicality score among the 3 fractionating methods and the mean preference ranking values for the 3 methods, the available data consistently supported the Time Method. The overall preferred and acceptable fractionating method for separating milk into foremilk and hindmilk, based on the evaluation of the mothers' preference (preference ranking), practicality score and willingness to perform the method for an extended time, is the Time Method. 68 Results 4.4 PHASE II - REPRODUCIBILITY OF A PROTOCOL FOR FRACTIONATING EXPRESSED HUMAN MILK INTO FOREMILK AND HINDMILK The Time Method was the preferred method for fractionating milk into foremilk and hindmilk. This method was used with a second group of mothers who delivered premature infants. Two criteria were used to determine the method's reproducibility: an increase in fat content from foremilk to hindmilk, where hindmilk contained at least a 1.5 fold higher fat content than foremilk, and the volume of expressed hindmilk collected was greater than the infant's prescribed enteral intake. Eight mothers were recruited and completed the baseline period in Phase II. Six of the 8 mothers successfully completed the fractionating period using the Time Method. 4.4.1 Fat Content of Preterm Milk The fat content of breast milk expressed by mothers of premature infants (preterm milk) collected during the baseline period in Phase II is shown in Table 4.15. The mean fat content of mothers' milk (n=8) was 3.65 ± 0.95 g/100 mL, as determined gravimetrically following extraction of the fat by the modified Folch method (Folch et al 1956). The range of fat content was 2.67- 4.99 g/100 mL. The mean creamatocrit value determined on fresh milk (n=8) was 6.53 ± 1.18 % and on frozen milk was 7.19 ± 1.43%. The corresponding mean fat content estimated, using the equation developed by Lucas et al. (1978), was 3.95 ± 0.92 g/100 mL for fresh milk and 4.52 ± 0.99 g/100 mL for frozen milk. The mean total fatty acid content of milk was 3.59 ± 1.04 g/100 mL. The mean T G content of milk was 3.77 ± 1 . 1 7 g/1 OOmL (TG content was determined by applying a factor of 1.05 to the fatty acid content of milk). The mean total fat content of milk determined by analysis of total fatty acid content using gas liquid chromatography and then converted to total fat content (by applying a factor of 1.063) was 3.82 ± 1.09 g/100 mL. 69 Results Table 4.15 Fat Content of Preterm Milk in Phase II1 Fat Analysis Method Mean ± SD Range n=8 Gravimetric Method * (g/100 mL) 3. 65 ± 0.95 2.64-4.99 Creamatocrit Method Fresh Milk3 (g/1 OOmL) 3.95 ± 0.92 2.82-5.30 Creamatocrit Method Frozen Milk4 (g/1 OOmL) 4.52 ± 0.99 3.22-5.88 Fatty Acid Method 5 (g/1 OOmL) 3.82 ± 1.09 2.23-5.35 (total fat content) Fat content was determined for milk samples collected during the baseline period in Phase II. 2 Fat extracted by the modified Folch Method (1956) and quantified gravimetrically. 3The Ceamatocrit value for fresh milk was 6.53 ± 1.18% (mean ± SD), range 4.83-8.30%. The fat content was calculated from the creamatocrit value using the Lucas Equation (1978): [fat (g/L)=(creamatocrit value -0.59)/0.146]. 4 The Creamatocrit value for frozen milk was 7.19 ± 1.42%(mean ± SD), range 5.30-9.17%. The fat content was calculated from the creamatocrit value using the Lucas Equation (1978): [fat (g/L)=(creamatocrit value - 0.59)/0.146] 5 Fatty acid content was determined by gas liquid chromatography and then converted to triacylglceride (TG) and total fat content by applying the conversion factor 1.05 & 1.063, respectively. The total fatty acid content of milk was 3.59 ± 1.04 g/100mL(mean ± SD), range 2.10-5.03 g/1 OOmL. The TG content was 3.77 + 1.17 g/1 OOmL (mean ± SD), range 2.21-5.28 g/100mL. 4.4.2 Fat content of Foremilk and Hindmilk Six mothers successfully completed Phase II by separating their expressed breast milk into foremilk and hindmilk the Time Method. The fat content of foremilk and hindmilk determined gravimetrically is summarized in Table 4.16. The mean fat content of foremilk was 2.36 ± 0.51 g/1 OOmL and the mean fat content of hindmilk was 4.80 ± 1.03 g/100 mL. The mean difference in fat content between foremilk and hindmilk was 2.44 ± 0.97 g/100 mL. The mean relative increase in fat content from foremilk to hindmilk was by a factor of 2.1 ± 0.53. Table 4.16 Fat Content (g/1 OOmL) of Foremilk and Hindmilk Using the Time Method1 Mean ± S D n=6 Range Foremilk (g fat/1 OOmL) 2.36 ± 0 . 5 1 1.94-3.30 Hindmilk (g fat/1 OOmL) 4.80 ± 1.03 3.58-6.22 Absolute Difference2 (g fat/1 OOmL) 2.44± 0.97 1.17-3.49 Relative Increase in fat content3 2.10 ± 0 . 5 3 1.50-2.80 Values are reported for mothers who successfully completed Phase II. ! The absolute difference refers to the differer 'The relative increase refers to the factor by \ (hindmilk fat content foremilk fat content). 2 rence in fat content from foremilk to hindmilk. 3 which the fat content of hindmilk was increased relative to foremilk 70 Results 4.4.3 Volume of Expressed Preterm Milk 4.4.3.1 Volume of Expressed Preterm Milk During the Baseline Period The volume of breast milk expressed by mothers during the 24 hour baseline period in Phase II is summarized in Table 4.17. The mean volume of milk expressed by mothers was 823 ± 384 mL per day (range 421-1355 mL, n=8). Three mothers (2 mothers who delivered singletons and 1 mother who delivered twins) expressed more than 1000 mL milk per day. All the mothers produced more than their own infant's prescribed enteral intake at the time of the baseline collection. The percentage of milk volume expressed by mothers relative to their infant's intake ranged from 138% to 1255% (n=8). Table 4.17 Volume of Expressed Preterm Milk During the Baseline Period1 Mean ± SD Range n=8 Milk volume (mL/day) 823 ±384 421-1355 Results are shown for milk volumes collected during the baseline period in Phase II. 4.4.3.2 Foremilk and Hindmilk Volumes Six mothers successfully completed Phase II by separating their expressed breast milk into foremilk and hindmilk using the Time Method. The volume of foremilk and hindmilk expressed using the Time Method is reported in Table 4.18. The mean volume of foremilk expressed was 339 ± 182 mL (range 150 - 580 mL, n=6). The mean volume of hindmilk expressed was 350 ± 167 mL (range 255 - 648 mL, n=6). The volume of hindmilk produced as a percentage of the total volume of milk expressed by the mothers (hindmilk volume divided by the total volume of milk expressed x 100%), ranged from 35-66%, with a mean of 52%,n=6. 71 Results Table 4.18 Foremilk and Hindmilk Volumes Expressed Using the Time Method1,2 Mean ± SD Range n=6 Foremilk (mL/day) 339 ± 1 8 2 150-580 Hindmilk (mL /day) 350 ± 1 6 7 255 - 648 Hindmilk relative to total milk volume 3(%) 52 ± 1 1 35-66 Hindmilk volume relative to infant intake 4 178 ± 129 97-438 Results are shown for mothers who successfully completed the milk fractionating period in Phase II. 2Six mothers completed 2 fractionating days and 2 mothers completed all 3 fractionating days. The data collected from the fractionating days was averaged for each mother, then used in the analysis. 3Hindmilk volume is reported as a percentage of total milk volume [(hindmilk volume-;- total volume) x 100%]. 4Hindmilk volume is reported as a percentage of infant intake [(hindmilk volume •*- infant intake) x 100%]. The number of mothers whose volume of hindmilk collected by the Time Method did not meet their own infant's enteral intake is illustrated in Figure 4.4. The percent of hindmilk volume expressed relative to the infant's prescribed intake [(hindmilk volume infant's intake) x 100%] ranged from 97-438%. Only 1 mother was unable to meet her infant's intake when fractionating hindmilk with the Time Method in Phase II. However, this volume of hindmilk was very close to the infant's intake (97% of intake). All the mothers expressed a volume of hindmilk within at least ± 5% of their own infant's enteral intake. 72 Results Figure 4.4 Ability of Volume of Mothers' Expressed Hindmilk to Meet Infants' Intake (Phase II) 1 Bars represent each mother's volume of hindmilk as a percentage of her own infant's intake [(hindmilk volume infant's intake) x 100%] 2 Results are shown for mothers who successfully completed Phase II, n=6. 73 Results 4.5 PHASE III - DETERMINATION OF VOLUME OF BREAST MILK EXPRESSED BY MOTHERS WHO DELIVER PREMATURE INFANTS 4.5.1 Volume of Expressed Preterm Milk Thirteen mothers in Phase III completed the breast milk expression diary. The volume of breast milk expressed by mothers in Phase III is summarized in Table 4.19. The mean volume of milk expressed was 641 ± 293 mL/day and ranged from 231-1288 mL. Two mothers (1 mother who delivered a singleton and 1 mother who delivered twins) expressed over 1000 mL milk per day. The number of mothers whose milk production was unable to meet their own infant's prescribed enteral intake is illustrated in Figure 4.5. Two of the 13 mothers were unable to meet their infant's intake (volume expressed <100% of infant's intake). Five of the 13 mothers were unable to express > 125% of their infant's intake, which was the minimum criteria, established for Phase I and II for fractionating milk. The percentage of milk volume expressed by mothers relative to their infant's intake ranged from 77% to 441%. Table 4.19 Volume of Expressed Preterm Milk in Phase III1,2 Mean ± SD Range n=13 Milk Volume (mL/day) 641 ± 293 231-1288 1 Results are shown for mothers who completed the breast milk expression diary in Phase III. 2 Twelve mothers completed a 3-day milk expression diary and 1 mother completed a 2-day milk expression diary. The data collected from the milk diary was averaged for each mother and then used in the analysis. 4.5.2 Expressed Breast Milk Practices Survey The results of the survey of expressed breast milk practices from Phase III are presented in Appendix 12 in combination with the survey results from Phase I and II. Fifteen of the 19 mothers in Phase III completed the survey. The expressed breast milk practices survey explored potential factors that may affect the mothers' milk expression routine and milk production. The results of the prevalence of these factors are presented in Appendix 12, as well as selected comments from mothers regarding factors they felt affected their milk supply, let-down reflex and milk expression routine. 74 Results Figure 4 . 5 Ability of Volume of Mothers' Expressed Milk to Meet Infants' Enteral Intake1'2 1 Each bar represents one mother's volume of expressed milk as a percentage of her own infant's intake [(hindmilk volume infant's intake) x 100%]. 2 Results are shown for mothers who completed the breast milk expression diary in Phase III, n=13. 75 Results 4.6 RELATONS BETWEEN THE MODIFIED FOLCH METHOD AND THE CREAMATOCRIT METHOD 4.6.1 Correlation between the Creamatocrit Value and the Total Fat Content of Expressed Breast Milk Expressed breast milk samples collected from mothers in Phase I and II were used to examine the relationship between the 2 methods of fat analysis, the modified Folch Method and the Creamatocrit Method. There was no significant difference in the creamatocrit value between fresh and frozen milk samples (Table 3.1), therefore results are provided for frozen milk samples only. One hundred and fifty-five frozen expressed breast milk samples were obtained from 29 mothers for the analysis. The milk samples consisted of baseline, foremilk and hindmilk samples and depicted a wide range of fat concentrations from 1.4 to 7.5 g/100 mL when determined by the modified Folch Method, and 3.0 to 12.7% when determined by the Creamatocrit Method. There was a highly significant correlation of the fat content of milk determined by the modified Folch Method and the creamatocrit value determined by the Creamatocrit Method ,r=0.953, P<0.001 ,n=155 (Figure 4.6). 76 Results 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Creamatocrit (%) Value Figure 4.6 Relations Between Fat Content Values as Determined by the Modified Folch Method and the Creamatocrit Method1'2 , 3 1 Results are show for frozen expressed breast milk samples collected from 29 mothers in Phase I & Phase II, n=155. 2Creamatocrit Value determined by the Creamatocrit Method; milk fat content (g/1 OOmL) determined by the modified Folch method. 3Solid lines represent the regression line (middle line) and 95% confidence interval (outer lines); r=0.953, P<0.001, regression equation: Y'=0.572X-0.180. 4.6.2 Correlation between Creamatocrit Value and Total Fat Content of Foremilk and Hindmilk Samples To further explore the relations between the fat content of milk determined by the 2 different methods, analyses were also performed on foremilk and hindmilk samples. Sixty-three frozen expressed milk samples were obtained from 25 mothers from Phase I and Phase II for the analysis. The foremilk samples represented a range of fat concentrations from 1.41 to 4.23 g/100 mL when determined by the modified Folch Method, and 3.0 to 8.5% when determined by the Creamatocrit Method. The hindmilk samples represented a range of fat concentrations from 2.90 to 7.49 g/100 mL when 77 Results determined by the modified Folch Method, and 5.5 to 12.7% when determined by the Creamatocrit Method. There was a highly significant correlation of the fat content of milk determined by the modified Folch Method and the creamatocrit value determined by the Creamatocrit Method for foremilk, r=0.944, P<0.001 ,n=63 (Figure 4.7a). There was also highly significant correlation of the fat content of milk determined by the modified Folch Method and the creamatocrit value determined by the Creamatocrit Method for hindmilk, r=0.875, P<0.001,n=63 (Figure 4.7b). 4.6.3 Development of a Regression Equation for Estimating Total Fat Content of Expressed Breast Milk Using the Creamatocrit Method Linear regression analysis was used to evaluate the relationship between the creamatocrit value determined by the Creamatocrit Method and the fat content of milk (g/1 OOmL) determined gravimetrically by the modified Folch Method. One hundred and fifty-five frozen expressed breast milk samples were obtained from 29 mothers for the analysis. A statistical significant correlation was found between the 2 measurements of fat content (g/1 OOmL and creamatocrit value) (Figure 4.6). A regression equation was developed to predict the estimated total fat content of milk from the creamatocrit value (Figure 4.6). The regression equation for the frozen milk samples is Y' = 0.572X - 0.180, (R2 =0.908, n=155), where Y' represents the milk fat content (g/1 OOmL) and "X' = the creamatocrit value. The regression equation for fresh milk samples (n=68) was similar to the regression equation developed for frozen milk samples (data not shown, see Appendix 8). The fat content of preterm milk can be predicted from a creamatocrit measurement (determined by the Creamatocrit Method) using the following equation: Fat (g/1 OOmL) = [(creamatocrit value) x 0.572] - 0.180. 78 Results a) Foremilk2 Creamatocrit (%) Value b) Hindmilk3 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Creamatocrit (%) Value Figure 4.7 Relations Between Fat Content Values as Determined by the Modified Folch Method and the Creamatocrit Method in Foremilk and Hindmilk Samples1 1 Results are show for frozen expressed breast milk samples collected from 25 mothers in Phase I & Phase II, n=63 2Solid lines represent the regression line (middle line) and 95% confidence interval (outer lines), r=0.944, P<0.001. 3Solid lines represent the regression line (middle line) and 95% confidence interval (outer lines); r=0.875, P<0.001. 79 Results 4.7 RELATIONS BETWEEN FAT CONTENT OF PRETERM MILK AND MILK VOLUME Expressed breast milk collected during the baseline period in Phase I and II were used to examine the relationship between milk fat content and milk volume. Twenty-seven mothers completed the baseline period, 19 mothers from Phase I and 8 mothers from Phase II. There was no relationship between the volume of milk produced and the fat content of the expressed breast milk. The scatterplot (Figure 4.8) illustrates this lack of a correlation between the 2 variables. E o o c Q) -t—i C o O •4—» CO o.u 5.0 4.0 H 3.0 2.0 • • —i • 1 1 1 1 1 « 1 • 1 1 1 — 200 400 600 800 1000 1200 1400 1600 Milk Volume (mL/day) Figure 4.8 Scatterplot of Breast Milk Fat Content and Milk Volume Produced by Mothers of Premature Infants1'2 Values represent 27 mothers who completed the baseline period in Phase I & II. 2 N o statistical significant relations were found, r=-0.034, P=0.864. 80 Results 4.8 SUMMARY OF RESULTS WITH RESPECT TO HYPOTHESES Hypothesis 1a: An acceptable method can be developed for separating the expressed milk of mothers who deliver premature infants into foremilk and hindmilk, which results in hindmilk with at least a 1.5 fold higher fat content than foremilk. All 3 fractionating methods (Visual Method, Time Method, and Volume Method) were able to meet the minimum criteria of at least a 1.5 fold increase in fat content from foremilk to hindmilk, thereby making all 3 methods acceptable based on the biochemical analysis criterion. Hypothesis 1 b: An acceptable method can be developed for separating the expressed milk of mothers who deliver premature infants into foremilk and hindmilk, in which the volume of hindmilk collected will meet the infant's prescribed enteral intake. The Visual Method and the Time Method were considered acceptable fractionating methods based on the hindmilk adequacy criterion, as most mothers were able to collect a sufficient quantity of hindmilk to meet their infants' prescribed enteral intake. Hypothesis 1c: A method for fractionating and collecting hindmilk from the expressed milk of mothers who deliver premature infants can be developed which is acceptable and practical for mothers to incorporate into their daily milk expression routine. The Time Method was considered the most acceptable fractionating method based on the practicality criterion. Although there was no significant differences found when comparing the mean practicality scores among the 3 fractionating methods and the mean preference ranking values among the 3 methods, the available data consistently supported the Time Method. 81 Discussion CHAPTER 5 DISCUSSION 5.1 Introduction A practical and reliable fractionating method for the selective collection of hindmilk is a key component of a hindmilk feeding strategy. This strategy utilizes a high fat, energy dense milk fraction to optimize the energy intake of premature infants, and thus promote better weight gain. Published research on fractionating expressed milk into foremilk and hindmilk for the purpose of feeding hindmilk to premature infants is limited. The present study was conducted to develop a protocol for the selective collection of hindmilk based on a fractionating method that is both practical and reliable. The first section of this chapter discusses the inter-individual variation of the fat content of preterm milk found in this study and the Creamatocrit Method, a simple and quick technique for estimating the fat content of milk. Next, the determination of an acceptable and reliable fractionating method for the selective collection of hindmilk is discussed. The criteria for evaluating the 3 fractionating methods involved the biochemical analysis of the fat content of the milk, the ability of the volume of hindmilk to meet the infant's prescribed enteral intake, and practical considerations for incorporating the method into the mother's daily pumping routine. The strengths and limitations of each method will also be discussed. The last section in this chapter discusses how the fractionating protocol could be applied in a clinical setting. This involves identifying potential candidates for the hindmilk fractionating protocol, using the Creamatocrit Method with the regression equation developed in this study to estimate the fat content of maternal milk, and developing guidelines for a fractionating protocol that can be implemented as part of a hindmilk feeding strategy. 5.2 The Fat Content of Preterm Milk Knowledge of the fat content of mother's milk (thus energy density of milk) would assist the clinician in assessing and evaluating the infant's energy intake. A simple and reliable method for estimating the fat content of milk in a clinical setting would help facilitate individualizing the infant's nutritional management. 82 Discussion 5.2.1 Inter-Individual Variation of the Fat Content of Milk The results from Phase I and II of this research show a wide range in the fat content of milk among mothers who deliver premature infants. Based on gravimetric analysis of the baseline milk samples (n=27), the mean fat content of mature preterm milk was 3.86 ± 0.86 g/1 OOmL, with a range of 2.33 to 5.40 g/1 OOmL (Appendix 9). The results of this study compare well with published values on the fat content of preterm milk. Anderson et al. (1981) reported a similar range in the fat content of preterm milk, approximately 2.7 - 5.5 g fat/1 OOmL, produced by 15 mothers at day 15-18 postpartum. The average fat content of mature preterm milk found in studies using a 24 hour collection of milk has been reported to be 3.5 - 4.5 g/1 OOmL (Anderson, 1981, Atkinson 1981, Ehrenkranz 1994, Lemons 1982). The wide inter-individual variation in the fat content of maternal milk found in this study and by others, may have clinical implications for infants who are born prematurely, especially if the milk fat content was consistently low. Typically, infants are fed a prescribed enteral volume of their own mother's milk. In clinical practice, infants fed expressed milk are assumed to be receiving milk with a fat content of 4g fat/1 OOmL, an energy content of 70 kcal/100mL, a protein content of 1.5 g/1 OOmL and a lactose content of 7g/100mL (Anderson 1981, Atkinson 1981, Gross 1984, Lemons 1982). These average composition values for mature preterm milk are used to calculate the infant's energy and nutrient intake. The wide variation in the fat content of milk among women suggests that some infants will receive milk with a lower fat content and subsequently a lower energy concentration than expected (assuming the protein and lactose content of milk is relatively constant). Infants receiving a low fat, low energy density milk would be at risk of suboptimal energy intake and poor growth. Based on the range of fat concentrations found in this study (2.33-5.40 g fat/1 OOmL milk), the corresponding energy density of milk would range from 56 - 84 kcal/100mL. This wide range in the energy density of milk suggests that some infants are receiving milk with lower energy content than expected, while others are receiving milk with higher energy content than expected. Of clinical importance is the infant who is receiving the low fat, low energy density milk. Twenty-five percent of mothers in the present study had a fat content of milk less than 3.26 g/1 OOmL, which corresponds to an estimated energy content less than 65 kcal /100mL (Appendix 9). Using the average fat (4 g fat/1 OOmL) and energy content (70 kcal/100mL) of preterm milk for these infants 83 Discussion would not accurately estimate their energy intake. Depending on the infant's enteral fluid intake, these infants would be at risk of subpotimal energy intake. Poor growth in these infants therefore may be a consequence of the mother's low milk fat content and low energy content. An example of the implications of varying fat concentration on the estimation of energy intake in a 1kg infant fed 150 mLVkg of unfortified milk and fortified milk is illustrated in Table 5.1. Infants of mothers who produce milk with a low fat, and thus low energy content, may not meet their energy requirements for growth when fed their own mother's milk. The estimated energy requirement for growth for preterm infants is >120-130 kcal/kg/day (AAP 1998, C P S 1995, E S P G N 1987). An infant receiving unfortified milk at 150 mL/kg/day would need to be fed milk containing at least 5.0 -5.5 g fat/1 OOmL to achieve an adequate caloric intake. An infant receiving fortified milk at 150 mL/kg/day would meet their estimated caloric requirement if their mother's milk fat content was at least 3.5 g /100mL. The energy density of mothers' milk was not determined in this study. This however, can be calculated using the following conversion factors based on the heats of combustion for the individual macronutrients: 4.4 kcal /g protein, 3.95 kcal/g carbohydrate and 9.25 kcal/g fat (Raiha 1999, Watts 1964). The gross energy content of milk estimated by this method will overestimate slightly the actual energy available to the infant as the heat of combustion values do not account for potential energy losses due to digestion or metabolism (Anderson 1981, Fomon 1993, Krause 1984,Watt 1964). Table 5.1 Comparison of the Estimated Energy Intake of a 1 kg Preterm Infant Fed Maternal Milk with Varying Concentrations of Fat1'2 Fat content of Maternal Milk (g fat/1 OOmL) 2.5 3 3.5 4 4.5 5 5.5 Energy intake of unfortified 86 93 100 107 113 120 127 milk (kcal/kg) Energy intake of fortified 107 114 120 128 134 141 148 milk3 (kcal/kg) Based on an intake of 150 mL/kg/day; energy content of milk calculated as the sum of 4.4 kcal/g protein, 3.95 kcal/g carbohydrate, and 9.25 kcal/g fat (Raiha 1999, Watts 1964); assuming preterm milk contains 1.5 g protein/1 OOmL and 7.0 g lactose/1 OOmL. 2The recommended energy intake for premature infants is > 120-130 kcal/kg/day (AAP1998, CPS 1995, ESPGN 1987). 3Preterm milk fortified with commercial human milk fortifier (Similac HMF, 4 packages added to 100mL; 4 packages contain 14 kcal). 84 Discussion Analysis of the fat and energy content of maternal milk would provide useful information for clinicians to identify mothers who produce milk low in fat and thus low in energy, which may place their infant at risk of marginal energy intakes. For premature infants who are growing poorly due to the low fat content of their mother's milk, feeding hindmilk (milk with a higher fat content) may be considered if the mother produces a sufficient volume of milk. 5.2.2 Determination of the Fat Content of Milk The determination of the fat content (and energy content) of maternal milk is necessary for assessing and evaluating the infant's actual energy intake. Therefore, in clinical practice it is important to have a reliable and practical method for estimating the fat content of milk. The following section discusses the methods used to analyze the fat content of human milk in this research. The 3 methods used in this study were a gravimetric method (modified Folch procedure), the Creamatocrit Method, and a fatty acid analysis method. The results of the total fat content of mature preterm milk from the baseline period of Phase I and II (n=27) showed a significant difference between the fat content determined by the gravimetric method (3.86 ± 0.86 g/1 OOmL; mean ± SD) and the Creamatocrit Method (4.52 ± 0.92 g/1 OOmL) (Appendix 9). The mean fat content of milk using the Creamatocrit Method with the equation published by Lucas et al. (1978) was 17% higher than the mean fat content of milk determined by the gravimetric method (modified Folch procedure). A possible explanation for this difference in fat content between the two methods might be due to potential losses of medium chain fatty acids, MCFA, during the lipid extraction procedure of the gravimetric method. MCFA are more soluble compared to longer chain fatty acids, therefore may be lost in the aqueous phase of the Folch procedure. The MCFA content of mature preterm milk (n=27) in this study was about 7% of the total fatty acid content of milk (Appendix 10). This result suggests that the MCFA content of milk alone did not account for the large difference found in the total fat content of milk determined by the 2 methods. The mean (± SD) fat content of milk determined by fatty acid analysis and using a conversion factor of 1.063 (to convert fatty acids to total milk fat) was 4.02 ± 0.90 g/1 OOmL. This result was not significantly different from the fat content of milk 85 Discussion determined by the gravimetric method. A small difference (4%) was found when comparing the fat content of milk determined by gravimetric method and the fat content of milk determined by the fatty acid method (fatty acid analysis + conversion factor). This small difference between the gravimetric and fatty acid analysis method may reflect a small loss of MCFA during the gravimetric method or it may reflect a slight overestimation when using the fatty acid analysis method. An overestimation using the fatty acid analysis method may be due to the assumptions made when converting fatty acids to T G and total fat content (e.g. assumptions regarding the amount of lipid in milk present as TG and phosholipids; the fatty acid methyl esters represent fatty acids from the TG molecules only; the molecular weight of common TG). However, these small differences suggest that the gravimetric method used in this study more accurately reflects the total fat content of milk than the Creamatocrit method, which uses the equation developed by Lucas et al. (1978). Another possible reason for the differences between the 2 methods could be related to fat concentration of milk samples used to develop the regression equation published by Lucas et al. (1978). The milk samples analyzed in Lucas's study contained low to average fat concentrations. More than half of the milk samples collected in their study were from "dripped" milk. "Dripped" milk is collected from the opposite breast during a breastfeeding session and tends to be low in milk fat (Gibson 1977). The equation developed by Lucas et al. (1978) for estimating the fat content of human milk, may not be applicable to the present data set that contains a wider range of milk fat concentrations, in particular milk with a high fat content. Based on the differences in the fat content of milk determined by the gravimetric method (modified Folch procedure) and Creamatocrit method (using the equation published by Lucas et al. 1978) it was necessary to develop a new regression equation for estimating the fat content of milk that would be more suitable for analyzing milk with varying fat concentrations. The Creamatocrit Method, using this new equation, could then be used in clinical practice as part of a fractionating protocol for estimating the fat content of maternal milk. 5.3 Determination of an Acceptable Fractionating Method The primary objective of this study was to determine an acceptable fractionating method for separating milk into foremilk and hindmilk. To do so, 3 fractionating methods 86 Discussion were compared and evaluated on several criteria, including the biochemical analysis of the fat content of milk, the ability of the expressed hindmilk volume to meet the infant's enteral intake, the subjective evaluation by mothers on the practicality of the method for incorporation into their daily pumping routine, the strengths and limitations of the methods and the reproducibility of the preferred fractionating method. 5.3.1. Fat Content of Foremilk and Hindmilk All 3 fractionating methods resulted in a significant increase in fat content from foremilk to hindmilk. This result is consistent with other studies that have examined the within feed variation of the fat content of human milk and have also reported a significant increase in the fat content of milk from foremilk to hindmilk (Dorea 1982, Hall 1979, Nelville 1984, Valentine 1994). All 3 fractionating methods met the minimum criterion of hindmilk containing at least a 1.5 fold increase in fat content from foremilk, thus making all 3 methods acceptable based on biochemical analysis. Valentine et al. (1994) studied the effect of feeding hindmilk (fortified with HMF) to15 low birth weight infants with poor weight gain. The investigators reported the mean (± SD) fat content of preterm foremilk and hindmilk was 2.86 ± 0.81 g/1 OOmL and 4.78 ± 0.85 g/1 OOmL, respectively. Milk was separated based on a specific time period of 2-3 minutes after "letdown". This fractionating method was the basis of the Time Method used in the present study. The results of the fat content of foremilk and hindmilk in the present study using the Time Method are comparable with the results published by Valentine et al. (1994). Slusher et al. (2000) studied the effect of hindmilk feeding to 18 low birth weight infants. The researchers provided no specific details on the fractionating method; however, the collection of hindmilk was based on the lipid content of milk using the Creamatocrit Method. The investigators reported the mean creamatocrit value for foremilk and hindmilk was 5% and 8.5%, respectively. These values are similar to the mean foremilk and hindmilk creamatocrit values obtained by the Visual Method (4.9% and 8.5%, respectively) and Time Method (5.1% and 8.8%, respectively) in the present study (Appendix 11). However the Volume Method had a higher mean creamatocrit value for foremilk (5.5%) and hindmilk (9.3%), than reported by Slusher et al. (2000). This difference in creamatocrit value may be due to differences in the fractionating 87 Discussion methods used to separate milk, the creamatocrit measurement tool used to determine the creamatocrit value, or the variability of fat content of milk among women. Vasan et al. (1998) studied the effect of individualizing the fat content of mother's milk on weight gain in extremely low birth weight infants. The creamatocrit technique was used to individualize hindmilk fractionation and hindmilk feeding. Results from their study showed the mean creamatocrit value for preterm hindmilk was 11.3%. This result is substantially higher than the creamatocrit value obtained in the present study by the Visual, Time or Volume Method. A possible explanation for this difference in the creamatocrit value may be related to an overall higher composite milk (foremilk + hindmilk) value of 8.4% reported in their study. In the present study the mean creamatocrit value for composite milk for the 3 fractionating methods ranged from 6.6-7.1%. Another possible reason for a higher creamatocrit value could be related to the point during milk expression when the hindmilk collection began. Vasan's data suggest that hindmilk was collected later during the milk expression session, as hindmilk collected later in a milk expression session would be expected to have a higher fat content. Although there was no significant difference in the fat content of hindmilk among the 3 methods used in the present study, a trend in the fat values was observed. The Visual Method resulted in the lowest mean hindmilk fat content compared to the Time and Volume Method. This result suggests that the Visual Method had a "point of separation" that occurred earlier in the pumping session than the other 2 methods. Fractionating milk early in the session would be expected to have a lower fat content than fractionating milk later in the session, because the fat content of milk increases during a feed or milk expression (Hytten 1954b). This result may be a consequence of the subjectivity of the Visual method, which relies on the mother's observation of the change in the appearance of her milk from thin and clear to thick and creamy. The Volume Method resulted in the highest mean hindmilk fat content compared to the other 2 methods. This suggests that the milk was separated later in the pumping session compared to the other 2 methods. Milk expressed at the end (or later) in the pumping session would have a higher fat content, as the fat content of milk increases throughout the session. This result was expected as the Volume Method is based on separating milk as late as possible in the pumping session that would allow for the collection of a sufficient volume of hindmilk to meet the infant's enteral needs. The 88 Discussion Volume Method may be more advantageous than the other 2 methods for collecting an energy dense hindmilk fraction because it resulted in the highest fat hindmilk content. 5.3.2 Volume of Foremilk and Hindmilk There were no significant differences in the volume of foremilk and hindmilk collected by the 3 fractionating methods. Despite no significant differences among the 3 fractionating methods, a trend in the data was observed for the Volume Method. The Volume Method resulted in a lower mean hindmilk volume compared to the other 2 methods. This lower hindmilk volume also corresponded to a higher fat content of milk indicating that the collection of hindmilk by the Volume Method occurred later in milk expression compared to the other 2 methods. The lower percent hindmilk volume (46%) resulting from the Volume Method may reflect the more precise nature of this method, which involves the collection of hindmilk volume that is intended to just meet the infant's enteral fluid prescription. The results on the percent of hindmilk volume relative to the total volume of milk expressed found in this study are different from results published in other studies involving fractionating maternal milk. Valentine et al. (1994) reported a mean volume of hindmilk that represented 60% of the total volume of milk expressed. In the research here, the mean hindmilk volume collected with the same method (Time Method) was 54% of the total volume of milk expressed. A possible explanation for the larger percent hindmilk volume reported in Valentine's study may be related to the duration of milk expression; mothers may have pumped for a longer time period than the mothers in the present study. Another possible explanation could be that mothers may have separated their milk at the 2 minute mark compared to the 3 minute mark from letdown which resulted in a lower foremilk volume and hence higher hindmilk volume. Slusher et al. (1999) reported a mean hindmilk volume representing 65% of the total volume of milk expressed. This result was higher than the percent hindmilk volume found in the present study. The higher percent hindmilk volume suggests that the point of separation (foremilk to hindmilk) occurred earlier during the milk expression session. The investigators reported that the creamatocrit was used to determine the point in the milk expression procedure when hindmilk collection began. However, no data were reported on what target creamatocrit value was used as the point for separating milk into 2 fractions. Slusher et al. (2000) reported a wide range in the volume of composite 89 Discussion milk (foremilk + hindmilk) produced by mothers of premature infants (200 -1020mL/day), which is similar to the ranges found in the present study. The Visual and Time Method were considered acceptable methods based on the collection of a sufficient volume of hindmilk to meet most of the infants' enteral intake. Only 1 mother (014) was unable to collect a sufficient hindmilk volume using the Visual Method. This mother reported she was not able to see a change in the color or consistency of her milk during milk expression. She picked an arbitrary point to separate her milk, for which the hindmilk volume represented 53% of the total volume of milk expressed. The problem experienced by this mother illustrates a potential limitation of the Visual Method, which relies on a subjective evaluation of the point when milk changes from foremilk to hindmilk. Two mothers were unable to collect enough hindmilk to meet their infant's enteral intake using the Time Method. One of these mothers (014) reported that the "majority of (her) milk was pumped within the first 3 minutes". Subsequently, she collected a very low quantity of hindmilk (35% of the total milk expressed), which did not meet her infant's enteral needs. This example illustrates a potential limitation of the Time Method. The rate of milk flow is an important factor to consider with the Time Method as some mothers who experience a quick flow of milk may express the majority of their milk in the first few minutes of expression. This leads to a collection of a lower volume of hindmilk and hence an inability to meet the infant's needs. An adjustment in the time designated for the collection of foremilk may be necessary for these mothers. For example, the "point of separation" could be adjusted to 2 minutes from "letdown" (or the start of pumping) instead of 3 minutes. The Volume Method was the least acceptable method based on the hindmilk volume criterion, as several mothers were unable to express a sufficient quantity of milk to meet their infant's enteral needs. One mother (04) produced 1065 mL during the baseline period. Based on her infant's intake, only 16% (170 mL/day) of her total milk volume was required as hindmilk to feed her infant. However, an insufficient volume of hindmilk was collected on the study day because the mother's volume of milk decreased to 850mL. Consequently, a much larger portion of milk was collected as foremilk, leaving only a smaller portion of milk to be expressed as hindmilk. A limitation of the Volume Method is that it assumes the volume of milk expressed is constant from day to day. The Volume Method is not sensitive to changes in the mother's milk production. Decreases in the mother's milk supply, in response to factors such as stress 90 Discussion or decreased pumping frequency, may result in an insufficient collection of hindmilk and thus an inability to meet the infant's enteral needs. 5.3.3 Practicality Evaluation of the Fractionating Methods Based on practical considerations, the Visual Method was the least acceptable fractionating method for mothers in this study. The willingness of the mother to perform the fractionating method for an extended period of time was felt to be a strong indicator of acceptability. If a mother was unwilling to perform a method, then a clinical protocol based on such a method would not be successful if implemented for the purpose of hindmilk feeding. The Visual Method had the most mothers (30%) who were not willing to use the method for an extended period of time compared to the other 2 methods. Two of these mothers reported they were unable to notice a change in their milk color or consistency when pumping. This factor may have resulted in their unwillingness to perform this method over a period of time. The convenience of the method may also be a factor in the mother's willingness to perform the method. The Visual Method was rated lower than the Time Method and similar to the Volume Method in the practicality score for the convenience criteria, however, these scores were not significantly different. Several mothers rated the Visual Method as "inconvenient" or "somewhat inconvenient". The Visual Method received a less favorable ranking of overall preference compared to the Time Method, with only 2 mothers ranking the method as "most preferred" and 4 mothers ranking the method as "least preferred". These results suggest the Visual Method was less desirable for many mothers to incorporate into their daily pumping routine. The Volume Method was less acceptable than the other 2 methods based on the mothers' preference and evaluation of practicality. Factors that limited the acceptability of the Volume Method were complexity and inconvenience. More mothers ranked the Volume Method as the "least preferred method" (50%). The mothers rated the convenience of the Volume Method lower than the Time Method. Several mothers rated the Volume Method as "somewhat convenient". One mother (013) who was unwilling to perform the Volume Method for an extended period of time commented that the method was "inconvenient" and "required good math skills" to perform. These factors may have explained why this mother reported that she would not perform this method for an extended time. The mothers also rated the Volume Method lower on the 91 Discussion practicality criteria for "clear and understandable directions" when compared with the other 2 methods. Although the majority (70%) of mothers were willing to perform the Volume Method over time, the practicality and overall preference of this method was less favorable when compared to the other 2 methods. The Time Method was considered an acceptable method based on practical considerations for incorporating the method into the mother's daily milk expression routine. Compared to the other 2 methods, the Time Method had the highest mean practicality score and more mothers ranked this method (rather than the other methods) as "most preferred". The Time Method also had the most mothers (80%) who were willing to perform the method for an extended period of time. Most mothers rated the Time Method higher on the "convenience" of the method compared with the other 2 methods. The mothers rated the Time and Visual Method similarly for "clear and understandable directions". All 3 methods were rated similarly for "time to complete the method". In summary, the mothers' evaluation of the practicality of the fractionating methods indicated the Time Method was the most preferred and practical for incorporation into their usual milk expression routine. The mothers' pumping routine, milk production, and milk characteristics (e.g. milk fiow and milk appearance) were factors that appeared to influence their evaluation and acceptance of the fractionating method. 5.3.4 Strengths and Limitations of the Fractionating Methods The strength and limitations of each method were identified based on the results of this study, the mothers' feedback and the researchers observations. The Visual Method is a simple method to perform, in that it does not require special equipment or complex instructions. In this method milk is separated based on the point when a change is noticed in the color and/or consistency of the milk expressed. Several mothers in the study, however, were unable to notice a change in the appearance of their milk during the milk expression session. These mothers therefore, used an arbitrary point during milk expression. Due to the subjective nature of the Visual Method, the "point of separation" chosen by the mothers could differ from session to session or from day to day. This potential variability could result in collection of inconsistent volumes of hindmilk, and consequently result in collection of milk of 92 Discussion differing fat content. In this research (Phase I), milk was collected for one 24 hour period. As a result, day to day variability of milk collected with this method could not be examined. Although the Visual Method is simple, this method may not be the most suitable for mothers who are unable to see a change in the appearance of their milk during milk expression. The basic concept of the Volume Method is to separate milk based on a predetermined volume of milk expressed at the beginning of milk expression. This method takes into consideration the volume of milk expressed by the mother relative to her infant's intake. The Volume Method is more complex than the other 2 methods because it involves more detailed instructions for the collection procedure. The volume of foremilk to be expressed at each pumping session is calculated, based on the total volume of milk expressed on a reference day, such that the remaining volume of hindmilk will meet the infant's enteral intake. The Volume Method involves more preparation for separating milk into foremilk and hindmilk because the mother must mark the collection bottle to indicate the predetermined foremilk volume to collect. A limitation of the Volume Method is that it is not sensitive to changes in the mother's milk supply. This method is not suitable for mothers whose milk production varies from day to day, or decreases significantly within a short period of time in response to factors such as stress or decreased pumping frequency. On the other hand, the Volume Method appears to be suitable for mothers who produce a large milk volume and whose milk production shows little fluctuation. The Time Method is an easy method to perform and requires only a watch or clock to assist in determining the "point of separation" for the collection of hindmilk. The basic concept of the Time Method involves separating milk based on a designated time (2-3 minutes) from the start of milk expression or letdown. One potential limitation of the Time Method is the rate of the mother's milk flow. The rate of milk flow is an important factor to consider when using the Time Method, as some mothers who experience a quick flow of milk may express the majority of their milk in the first few minutes of pumping. This leads to collection of a lower volume of hindmilk and hence inability to meet the infant's needs. An adjustment in the time from letdown (or start of pumping), which is used to indicate the point of separation of milk into foremilk and hindmilk, would be necessary for these mothers. For example, the "point of separation" could be adjusted to 2 minutes from letdown (or the start of pumping) instead of 3 minutes. 93 Discussion 5.3.5 Summary of Results: Determination of an Acceptable Fractionating Method The Time Method was chosen as the most acceptable and practical fractionating method for the selective collection of hindmilk in Phase I of this study. Although no significant differences were found when comparing the three fractionating methods based on the biochemical analysis of the fat content of milk (foremilk, hindmilk, difference from foremilk to hindmilk), the volume of foremilk and hindmilk produced or the practicality evaluation by the mothers (practicality score, preference ranking), the available data consistently supported the Time Method. Feedback from the mothers, trends in the data, observations made during the study, and the strengths and limitations of the methods provided useful information to help determine the most acceptable and practical fractionating method for the collection of hindmilk for mothers who deliver premature infants. The hindmilk volume data indicated that the Time and Visual Method were both acceptable methods for collecting an adequate volume of hindmilk to meet the infant's enteral intake. However, several mothers indicated difficulty in determining the point of separation (from foremilk to hindmilk) when using the Visual Method due to an inability to differentiate a change in their milk's appearance (milk color and consistency), and therefore chose an arbitrary point of separation. The highly subjective aspect of the Visual Method was felt to be a strong limitation of this method. Also, more mothers indicated an unwillingness to perform the Visual Method for an extended period of time, compared to the other fractionating methods. The Visual Method was considered a less desirable fractionating method compared to the Time Method, based on these factors and the mothers' comments. Factors that indicated the Time Method was the most preferred and more practical fractionating method for incorporating into the mothers usual milk expression routine were: the trends in the results of the preference ranking evaluation, the mothers comments and feedback regarding the method, and finding only a minor limitation of the method (compared to the limitations found with the other methods). The Time Method appeared to have a minor limitation, in that mothers with a rapid milk flow could produce most of their milk in the first few minutes of pumping. An adjustment to the designated time period for the collection of foremilk can be easily made to correct for this factor. The Time Method had the highest number of mothers (60%) ranking the method as the most preferred method. Several mothers commented that the Time Method was "easy" 94 Discussion to perform. This may be related to the limited equipment required (equipment required was a clock or watch) for separating milk, the limited preparation or set up required (no marking of bottles required to determine point of separation), and the elimination of the decision making process (lack of subjective component to fractionating) for determining when to separate milk into foremilk and hindmilk, hence "less thinking" required. The Visual Method may be appropriate as a fractionating method in some circumstances, such as, when the mother can clearly identify changes in her milk characteristics. However, the reliability of the Visual Method was not investigated in this study and may be warranted (due to the subject nature of this method) to determine if the method can result in an adequate change in fat content from foremilk to hindmilk (minimum 1.5 fold increase in fat content) and result in an adequate volume of hindmilk to meet the infant's enteral intake, especially if the method was to be used for an extended period of time. In summary, the Time Method could be used as a standardized method for separating milk of mothers with premature infants in the S C N . Mothers preferred to separate their milk into foremilk and hindmilk using this method due to the ease and convenience of the method. Minor adjustments to the Time Method may be indicated for individualizing the method for the mother's pumping routine, milk characteristics or milk production. For example, a mother with a rapid milk flow can use a 2-minute mark for separating her milk into foremilk and hindmilk. A mother who pumps for a long period of time with the single pumping technique (i.e. greater than 20 minutes) could use the 3-minute mark for separating her milk. A mother with a large volume of milk relative to her infant's intake could also use the 3-minute mark for separating her milk into foremilk and hindmilk. It is also possible to utilize various aspects of the Visual or Volume Method to make adjustments to the Time Method, if the situation were appropriate. To test the reproducibility of the Time Method, the most preferred and practical fractionating method, a different group of mothers from the S C N were studied and asked to collect hindmilk using this method for 2-3 non-consecutive days. 5.3.6 Reproducibility of the Time Method in Phase II The results of the Time Method in Phase II showed that the method was reproducible with a second independent group of mothers of premature infants. The Time Method in Phase II was able to meet the minimum criteria for a change in fat 95 Discussion content of milk from foremilk to hindmilk. All 8 mothers were able to demonstrate at least a 1.5 fold increase in fat content from foremilk to hindmilk, which was the objective. The majority of mothers in Phase II were able to express a sufficient quantity of hindmilk to meet the enteral intake of their infants. Only 1 mother (V01) did not collect a sufficient volume of hindmilk to meet her infant's enteral needs. However, this hindmilk volume was 97% of the infant's required intake and was considered acceptable for supporting the criteria for adequate hindmilk volume. 5.4 Clinical Application of the Fractionating Protocol A fractionating protocol for the selective collection of hindmilk can be used in a clinical and research setting as the basis of a hindmilk feeding strategy. The practical application of the protocol relies on the following conditions: (1) knowledge of the mother's milk production relative to her infant's intake; (2) having a simple and reliable method for determining the fat content of milk in the clinical setting; and (3) developing practical guidelines for implementing the fractionating protocol, thus the hindmilk feeding strategy. 5.4.1 Milk Production by Mothers Who Deliver Premature Infants: Considerations for Hindmilk Feeding The objective of Phase III was to determine what proportion of mothers participating in this research would be suitable candidates for fractionating their milk for the selective collection of hindmilk. Potential candidates for a hindmilk feeding strategy must express a sufficient quantity of milk relative to their infant's prescribed enteral intake to allow for fractionation. A sufficient quantity of milk is defined as a maternal milk volume greater than 125% of the infant's enteral intake. The results of this study (all 3 study phases) showed a wide range (231 -1435 mL/day) in the milk production by mothers who deliver premature infants. This wide variation of milk production among women is similar to results reported by other investigators (Hopkinson 1988, Slusher 2000, Lemons 1982, Anderson 1981, Hurst 1999). The mean (± SD) volume of mature preterm milk from Phase l/ll (n=27) and Phase III (n=13) were 749 ± 3 1 8 mL/day and 641 ± 293 mL/day, respectively. The milk production of mothers in Phase III was lower than Phase l/ll. This difference may be related to the wide variability in milk production among mothers, the smaller sample size 96 Discussion of Phase III compared to Phase l/lI, and/or the recruitment criteria for subjects participating in the study. The mothers participating in Phase I and II in this research were required to produce a sufficient quantity of milk (>125% of their infant's intake) to allow for fractionating into foremilk and hindmilk. These mothers were more likely to have a higher milk production than mothers recruited for Phase III. There were no volume requirements for mothers recruited for Phase III. Results from the present study (Phase I, II and III) showed a mean pumping frequency of 5-6 expressions per day and a mean pumping duration of 120 to 156 minutes per day. These results indicate that the mothers in this study were generally meeting the minimum recommendations for maintaining optimal milk production, as described by Hopkinson et al. (1988). The results of Phase III indicated that 8 of the 13 mothers (62%) produced a sufficient volume of milk to fractionate for the selective collection of hindmilk. Despite the small number of mothers participating in this study, these results suggest that a proportion of mothers may be suitable candidates for a hindmilk feeding strategy. However, the ability to generalize these results to the SCN population is limited and based on the results of Phase III it is difficult to assess what proportion of mothers in the SCN would be suitable candidates for the hindmilk fractionating protocol. Exploratory data was collected in this study to examine potential factors that may affect the mothers' milk production and milk expression routine (Appendix 12). Stress and a "busy schedule" were prominent factors identified by many mothers as affecting their milk production. Mothers who smoked cigarettes (n=7) were able to produce a sufficient volume of milk to meet their infants' enteral intake (all the mothers produced over 500ml_ per day) and this suggests that smoking did not affect their milk production. The milk produced by these mothers was higher than results reported by Hopkinson et al. (1992) who reported a mean milk volume by mothers who smoked cigarettes of 358 mL/day at 4 weeks postpartum. These investigators reported the milk production of mothers (of premature infants) who smoked cigarettes was significantly lower than the milk production of mothers who did not smoke cigarettes (Hopkinson 1992). It is difficult to interpret the results of the present study due the small sample size and the wide variations of milk production. Further studies are needed to explore factors that affect milk production in mothers who deliver premature infants. 97 Discussion 5.4.2 Creamatocrit Method as a Clinical Tool in the NICU The Creamatocrit method has been validated by several researchers (Lucas 1978, Mierer 1999, Lemons 1980, Wang 1999) and is considered a reliable technique for estimating the fat and energy content of human milk (Jensen 1989). The present study calibrated the creamatocrit value against the fat content of milk (g/1 OOmL) determined gravimetrically using a standard lipid extraction method. The results from data collected in Phase I and II (n=155 frozen samples) showed the creamatocrit value was strongly correlated with the fat content of milk (g/1 OOmL) determined gravimetrically (r =0.953, P<0.001). This result is similar to findings observed by Lucas et al. (1987), Wang et al. (1999), Meier et al. (1999) and Prentice et al. (1981). There are several published equations for converting the creamtocrit value (%) to total fat content of milk (g/1 OOmL). The regression equation developed in this study differs from other published equations, in that, it is based on a larger number of preterm milk samples, with a wider range of milk fat concentrations. A large number of milk samples (n=155) were used in the present research to establish the linear regression equation. The milk samples from Phase I and II represented a wide range of fat values, ranging from 1.4 to 7.5 g fat/1 OOmL as determined gravimetrically, and 3.0 -12.7% (creamtocrit value) as determined by the Creamatocrit Method. Lucas et al. (1978) were the first to describe the Creamatocrit Method as a technique for estimating the total fat and energy content of human milk. The regression equation reported by Lucas et al. (1978) appeared to be based on a high number of milk samples with low to average fat concentrations (approximately <4 g fat/1 OOmL and <6.5% creamatocrit value). This may be due to the high number of "dripped" milk samples analyzed in the study. The regression equation reported by Lucas et al. (1978) may not be suitable for predicting the fat content of milk with high fat concentrations, such as that typically seen in hindmilk samples. Wang et al. (1999) analyzed a small number of milk samples (n=17) and found a strong linear correlation between the creamatocrit value and the lipid concentration (g/1 OOmL) of preterm milk determined by gas chromatography, r=0.82, p<0.001 for frozen samples and r=0.92, p<0.001 for fresh samples. The creamatocrit measurements ranged from 2.0 to 7.9% and the lipid concentration of milk ranged from 1.1 to 4.4 g/1 OOmL. The range of fat measurements by the 2 methods was limited and consisted mostly of milk with low to average fat concentrations. The fat values obtained in Wang's 98 Discussion study were similar to the range of fat concentrations observed for the foremilk samples obtained in the present study (Phase I and II). Thus, the estimation of milk fat content based on the equation published by Wang et al. (1999) may not be appropriate for predicting the fat content of milk with high fat concentrations. Meier et al. (1999) reported a strong linear relation (r=0.94, p<0.001) between the creamatocrit value and the lipid concentration of maternal milk, determined by a gravimetric method. Eighteen mothers of premature infants provided 32 milk samples for analyses, consisting of 7 foremilk samples, 12 hindmilk samples and 13 composite milk samples (foremilk + hindmilk). The range of creamatocrit values was 5 .0 to 17.5% and lipid concentration was 2.88 to 8.65 g/1 OOmL. Despite the small number of samples analyzed, the results of their study are similar to the present study, in that a wider range of preterm milk fat contents (creamatocrit values and g fat/100 mL) were used to calibrate the Creamtocrit Method with a gravimetric method to produce a regression equation for estimating the total fat content of milk. Samples of foremilk (n=63) and hindmilk (n=63) were analyzed separately to allow further examination of the relation between the creamatocrit measure and the total fat content of milk determined gravimetrically. There continued to be a strong correlation between the two measurements of fat when examining foremilk (r=0.994) and hindmilk samples (r=0.875) separately. The strong relations with the foremilk and hindmilk samples indicate that the regression equation established in the present study, from all the samples (n=155), appears to be an acceptable predictor of the total fat content of milk for samples with both low and high fat contents. Appendix 13 compares the regression equation developed in the present study, for the estimation of total fat content of milk from the creamatocrit value, to other published equations discussed in this section. Table A9 shows the differences in the total fat content of milk (g fat/1 OOmL) derived by the different equations when varying creamatocrit values are used. This indicates the importance of calibrating the Creamatocrit Method with a standard lipid extraction method for an individual laboratory, as well as for the target population for which it will be used, i.e. preterm milk samples containing a wide range in fat concentrations. Having knowledge about an individual mother's milk fat content (and thus energy density of milk) by using the Creamtocrit Method, could aid clinicians in: (1) calculating a more realistic estimation of the energy intake of the infant; (2) identifying a 99 Discussion mother who produces a milk with a low fat content (and hence lower energy content), which may place their infant at risk of receiving a suboptimal energy intake; or (3) individualizing the infant's enteral feed with respect to fortification, supplementation, or implementation of hindmilk feedings. 5.4.3 Fractionating Protocol for the Selective Collection of Hindmilk A fractionating protocol based on the results and observations of this study can be used in a clinical setting as the basis for a hindmilk feeding strategy. The fractionating protocol developed in this study is based on the "Time Method" which was determined to be an acceptable and reliable method for the selective collection of hindmilk. This protocol utilizes the Creamatocrit Method, with the regression equation developed in this study, for estimating the total fat content of milk in a clinical setting. Suitable candidates for this fractionating protocol must produce a sufficient volume of milk relative to their infant's enteral intake to allow for fractionation. Box 5.1 describes the SCN fractionating protocol and guidelines for the selective collection of hindmilk that could be used in clinical practice and in future research. 5.4.4 Implications of Practice The Time Method was considered the most acceptable and practical fractionating method for the selective collection of hindmilk, an energy dense milk fraction. This fractionating method can be used as the basis of a hindmilk feeding strategy to improve the calorie intake and the weight gain of select premature infants. The decision to use the hindmilk feeding strategy as a nutritional intervention for premature infants fed mother's milk and who are growing poorly should be based on a thorough nutritional assessment. Several factors should be assessed to help determine the appropriateness of the hindmilk nutritional strategy for the individual infant. Some of these factors include: the infant's actual energy and nutrient intake (especially the protein intake), the infant's estimated requirements, the infant's growth status, the infant's feeding history, the infant's feeding tolerance, the infant's medical condition and the infant's prescribed total fluid intake. Infants identified with poor weight gain (defined as growth < 15g/kg/day) who have been receiving full enteral feeds of expressed breast milk fortified with a commercial human milk fortifier and who are receiving an adequate protein intake (3-4 g/kg/day) may be considered for the hindmilk nutritional strategy. However, the 100 Discussion Box 5.1 SCN FRACTIONATING PROTOCOL Guidelines for the Selective Collection of Hindmilk Background: Fat is a major source of energy in human milk, representing approximately 40-50% of the total calories. Fat is also the most variable macronutrient in human milk. It varies within a feed (from foremilk to hindmilk), during the day, with stage of lactation, and among women. The fat content of foremilk, milk at the beginning of a feed, has been reported to be 1.5 -3 fold lower than milk at the end of a feed (hindmilk). For some infants who grow poorly due to illness, high energy requirements, fluid restrictions, or prematurity, feeding hindmilk which is higher in fat content and hence calories, may be beneficial for promoting weight gain. Maternal milk can be separated into two fractions, foremilk and hindmilk, using a fractionating method based on time (2-3 minutes after the start of pumping or letdown). Research conducted in the SCN has shown that the "Time Method" is a reliable and practical method for the selective collection of hindmilk. Indications for Using the Fractionating Protocol: The fractionating protocol is intended for mothers whose premature infants have been placed on "hindmilk feedings" as a nutritional strategy to improve weight gain. The following conditions are required before implementation of the fractionating protocol (and hence hindmilk feeding strategy): 1) The premature infant is receiving full enteral feeds of fortified expressed breast milk (EBM + HMF). 2) The mother must be producing a sufficient volume of expressed milk, designated as >125% of their infant's enteral intake (see below). 3) The SCN Dietitian has been consulted for a nutritional assessment of the infant prior to the implementation of hindmilk feedings. Instructions: 1) A 24 hour milk expression record is required to determine the volume of milk expressed by the mother. If the mother produces >125% of her infant's enteral intake, the fractionating protocol can be implemented. 2) The mother is instructed to separate her milk into foremilk and hindmilk using the "Time Method". The "point of separation" will be 2-3 minutes after the start of pumping or 2-3 minutes after "letdown" (increased milk flow) has occurred. FO R E MILK HINDMILK 2-3 mins L E T D O W N Start of Pumping Point of separation (change gradufeed) 101 Discussion Box 5.1 (cont'd) 3) Mothers will follow the usual S C N guidelines for set-up, handling, labeling, and storage of expressed milk. Mothers will also label bottles as "hindmilk" and "foremilk". 4) Mothers will collect a small sample (1-2 mL) hindmilk (and foremilk if desired) from each container (gradufeed) pumped for a 24 hour collection period for analysis of fat content. Hindmilk samples collected for the 24 hour period will be pooled prior to analysis. Analysis of the Fat Content of Maternal Milk: The Creamatocrit Method is a quick and simple method for estimating the fat content of human milk. It requires a very small volume (~ 75 uL) of milk for analysis. The method requires a hematocrit centrifuge, hematocrit glass capillary tubes and a hematocrit reader (to determine the creamatocrit value). A well-mixed sample of milk is drawn up into a capillary tube. The tube is then sealed at one end and the sample is spun in a hematocrit centrifuge for about 10 minutes. Centrifugation of the milk sample results in the separation of the "cream" into a discrete layer. The height of the cream or fat layer is measured and expressed as a percentage of the total length of the milk column. This percentage is referred to as the "creamatocrit" value. The creamatocrit value can be converted to the total fat content of milk (g fat/1 OOmL) by using a special equation developed for the S C N . The following equation can be used to calculate the fat content of milk: Milk fat (g/1 OOmL) = [0.572 x (creamatocrit value)] - 0.180. The following table provides an easy reference guide for converting creamatocrit values to total fat values (g/1 OOmL). The average fat content of mature milk is 3.5-4.5 g/1 OOmL. Creamatocrit Total Milk Fat Content Value (%) (g fat/1 OOmL) 3 1.54 4 2.11 5 2.68 6 3.25 7 3.82 8 4.40 9 4.97 10 5.54 11 6.11 12 6.68 13 7.26 14 7.85 15 8.40 16 8.97 Please contact the S C N Dietitian for arranging the analysis of the milk sample. 102 Discussion mothers of these infants must produce a sufficient quantity of milk (>125% their infant's prescribed enteral intake) to allow for fractionation and hence the collection of hindmilk. The Creamatocrit Method, using the new regression equation developed in this study, can be used as a simple and quick clinical tool to estimate the fat content of the mother's milk (and hence energy content of milk). This information can be used to better estimate the infant's actual energy intake and to help assess the nutritional adequacy of the infant's diet. The Creamatocrit Method can also be used to identify mothers who produce milk with low fat content and subsequently low energy content. Poor growth in some infants may be due to the low energy concentration of their mother's milk. In these infants the hindmilk nutritional strategy may be indicated to optimize the infant's energy intake and thus promote weight gain (Anderson 2002). Premature infants who may be considered for this nutritional strategy are infants who are growing poorly and receiving a subpotimal energy intake most likely due to fluid restrictions, increased energy requirements, or the low energy density of their mother's milk. The hindmilk feeding strategy may be indicated as an alternative method to increasing the infant's energy intake without the supplementation or addition of term infant formula powder, polycose powder, MCToil or a hypercaloric infant formula (Kristen 1999). The hindmilk feeding strategy may not be indicated for infants who are growing poorly due to an inadequate protein intake. A protein supplement is warranted for these infants to optimize the protein intake to better meet their needs and promote adequate growth (Anderson 2002). Infants with fat malabsorption due Gl problems (e.g. short bowel syndrome) or liver problems (e.g. choleostasis), where digestion and absorption of fat are compromised, would not benefit from a nutrition strategy that utilizes a higher fat intake consisting primarily of long chain fatty acids. The use of hindmilk feedings in infants with chronic lung disease (i.e. bronchopulmonary dysplasia, BPD) would need further considerations to determine if the increased fat intake would lead to an undesirable increase in adipose tissue. For these infants, it is important to ensure both protein intake and energy intake are adequate to meet estimated needs for growth (Newkirk 1999). It is difficult to say if the hindmilk feeding strategy would be the most suitable nutrition intervention for infants with BPD, future research is warranted with this population. 103 Discussion When increasing the energy density of feeds, either by the use of hindmilk, or the addition of term infant formula powder or individual macronutrients (MCT oil, polycose powder), it is important that the distribution of calories from the macronutrients remain well balanced. The recommended calorie distribution of enteral feeds for infants are as followed: 9-12% of the total calories should come from protein, 40-45% of the total calories from carbohydrate, and 40-50% of the total calories from fat (fat should not be >60% of the total calories as this may lead to ketosis) (Fulhan2000, Newkirk 1999). Therefore, when implementing the hindmilk feeding strategy, having knowledge of the fat content of hindmilk (by using the Creamatocrit Method) will help determine the percentage of calories from fat. 5.5 Limitations of the Study A limitation in this study was the small number of the subjects participating in the study phases. A small sample size affects the power of the study and a low power has an increased probability of making a Type II error, i.e. the study fails to detect a true difference or effect. Post-hoc analysis was performed using the results obtained in the study to determine the number of subjects required to have an 80% probability of detecting a significant difference between the mean values of the dependent variable. The following equation was used to calculate the sample size with an a level of 0.05 and a (3 level of 0.20 (power 80%): n = [(Za+ Z p ) 2 x SD 2] /d 2; where n is the sample size, Za is the z value (1.96) corresponding to a=0.05; Zp is the z value (0.84) corresponding to (3 =0.20; SD is the standard deviation obtained from the study results; and d\s the difference between means (highest mean value - lowest mean value) (Norman 1994, Khachatryan 2000, Hazzard 1991). Using the standard deviations obtained in the study and aiming for a "medium" effect size, which is a treatment difference (difference between means) that is based on half the standard deviation (Hazzard 1991, Cohen 1977), the number of participants required to have an 80% probability of finding a significant difference or effect (or 20% risk of a Type II error) would be approximately 30. The recruitment of subjects in this study was a challenge. Difficulties in recruitment and maintaining participants in the study may be inherent to a high stress environment, such as that of a neonatal intensive care unit and the situation faced by the mothers, that of delivering a premature infant. Of the mothers identified as potential 104 Discussion subjects in Phase I and II, less than half were willing to participate. In Phase I, of the mothers who agreed to participate in the study less than half of these mothers were able to complete the study. Another limitation of the study was that the characteristics of the mothers participating in this study were fairly homogeneous. The participants were mostly Caucasian women, well educated (completed a post secondary program), highly motivated to participate in the study, and spoke and understood English well. These maternal characteristics may not reflect the overall SCN population or other neonatal populations. Therefore, the ability to generalize the results of this study may be limited. Thus, further research with a larger sample size and a more heterogeneous population (e.g. mothers with different ethnic backgrounds, different levels of education, and who may not speak English as their first language) may be warranted. 105 Conclusions and Future Directions CHAPTER 6 CONCLUSIONS AND FUTURE DIRECTIONS 6.1 Conclusions Major findings from this study are: (1) the Time Method is the most acceptable and practical fractionating method for the selective collection of hindmilk when compared with the Visual and Volume Method, (2) The Time Method is a reliable and reproducible fractionating method, (3) the Creamatocrit Method can be used for estimating the fat content of human milk as part of a fractionating protocol in a clinical or research setting, however, a regression equation suitable for predicting the fat content of milk with varying concentrations of fat is needed, and (4) the information from this research suggests a proportion of mothers of prematurely born infants are able to produce a sufficient quantity of milk, defined as >125% of their infant's enteral intake, to successfully fractionate and collect hindmilk, however, further studies with a larger and more heterogeneous group of mothers is needed before a conclusion can be drawn. 6.2 Future Directions This fractionating protocol could be used in clinical practice as a part of a hindmilk feeding strategy for premature infants with poor weight gain, however, prior to implementing this protocol as a routine feeding strategy in the SCN, further research is indicated. 6.2.1 Future Directions in Research 1) Further research on the milk production by mothers who deliver premature infants is needed to determine what proportion of mothers in the S C N would be potential candidates for the fractionating protocol. Due to the small number of subjects recruited in the present study, it was difficult to determine this information from the study results. A longitudinal study may be better for obtaining information on the mother's milk production relative to the infant's enteral intake at given time periods during the infant's hospitalization, for example, at 2 weeks, 4 weeks, 6 weeks and 8 weeks postpartum. 2) A wide variation in the fat content of milk among mothers was found in the study, suggesting that some infants would be receiving milk containing a low fat concentration, and hence a low energy content. This low energy dense milk may place the infant at risk for poor growth due to a suboptimal energy intake. To explore this potential problem, a 106 Conclusions and Future Directions study can be conducted to examine the relationship between infant growth and mother's milk fat content (energy content), to see if poor growth in some infants is related to the fat concentration of their mother's milk. 2) There is limited information on the nutrient composition of foremilk and hindmilk; most information generally pertains to the fat and energy content of the milk fractions. Valentine et al. (1994) provided data on the composition of nitrogen, fat, energy and a select number of minerals in foremilk and hindmilk. They did not study the fat-soluble vitamin concentrations in the milk fractions. The Vitamin A content of hindmilk has been reported to be 2-3 fold higher than foremilk (Valentine 1994). It would be expected that other fat-soluble vitamin concentrations would also be higher in hindmilk (higher fat milk fraction). More information is needed on the energy, macronutrient and micronutrient composition of hindmilk. This information will assist the clinician in individualizing the infant's nutritional management with respect to assessing the infant's nutrient intake, identifying potential nutrient deficiencies or excesses and determining the need for fortification or supplementation. Bomb calorimetry can be used to determine the energy content of hindmilk. The data from bomb calorimetry analysis (energy content of milk) could be used in conjunction with the Creamatocrit Method (creamatocrit value) to develop a regression equation to predict the energy content of milk from the creamatocrit value. This equation would be useful for clinicians when estimating the energy intake of infants who are fed human milk. 3) An intervention study is needed to examine the effect of feeding hindmilk, collected using the fractionating protocol developed in this study (based on the Time Method), to very low birth weight infants with poor weight gain (defined as <15 g/kg/day). The following outcomes could be considered for an intervention study involving feeding hindmilk: examination of the macronutrient and energy content of hindmilk to allow for the determination of the percent distribution of calories from fat, protein and carbohydrate (the Creamatocrit Method and bomb calorimetry could be used to determine the fat and energy content of milk); assessment of body composition of infants fed hindmilk (e.g. skinfold measurements (if possible), midarm circumference measures); assessment of the infant's growth status (weight, length, knee to heel length, head circumference measurements); assessment of feeding tolerance and fat 107 Conclusions and Future Directions absorption (fecal fat test); assessment of nitrogen retention; assessment of the feeding delivery system (bolus feeds versus continuous feeds) when feeding the higher fat milk fractions, as the adherence of fat to the feeding tubes may result in a significant loss of fat and hence energy (Jones Wessel 2000); and examination of the safety of feeding hindmilk for extended periods of time. 4) Further research is indicated using the acceptable fractionating method determined in this study, the Time Method, with a more heterogeneous group of mothers to see if similar results can be obtained with mothers of a different background (e.g. teenage mothers, mothers with limited English speaking ability, and mothers with a lower education level). Also, a future study may be indicated to examine the reliability of the Visual Method to see if this method, when performed over a period of time, can result in an acceptable change in fat content of milk from foremilk to hindmilk, as well as, result in the collection of an adequate volume of hindmilk to meet infant's intake. 6.2.2 Future Directions in Practice 1) Information obtained in this study on the mother's milk production will be shared with members of the neonatal team (e.g. nurses, lactation consultants, doctors) to help support and encourage lactation in mothers who deliver premature infants. 2) The Time Method could be used as part of a hindmilk feeding strategy to increase the energy intake of infants identified with poor growth. Future research on hindmilk feedings is indicated prior to the establishment of this feeding strategy as a routine clinical intervention for infants with suboptimal energy intake and poor weight gain. 3) The Creamatocrit Method is a simple and quick technique that can be used in a clinical setting to estimate the fat content of mother's milk, and hence the energy content of milk. This information will better assist clinicians in assessing the energy intakes of infants fed their own mother's milk, identifying mothers who produce milk with a low fat content (hence low energy content) which may place their infant at risk of suboptimal energy intake, and individualizing the infant's enteral feeds with respect to fortification, supplementation or implementation of hindmilk feedings. 108 References REFERENCES American Academy of Pediatrics Committee on Nutrition. (1977). Nutritional needs of low-birth-weight infants. Pediatrics, 60, 519-530. American Academy of Pediatrics Committee on Nutrition. (1985). Nutritional needs of low-birth-weight infants. Pediatrics, 75, 976-986. American Academy of Pediatrics Work Group on Breastfeeding. (1997). Breastfeeding and the use of human milk. Pediatrics, 100, 1035. American Academy of Pediatrics Committee on Nutrition. (1998). Nutritional needs of the preterm infants. In R.E. Kleinman (Ed.). Pediatric nutrition handbook (pp. 55-87).llinois: American Academy of Pediatrics. Anderson, G.H., Atkinson, S.A., Bryan, M.H. (1981). Energy and macronutrient content of human milk during early lactation from mothers giving birth prematurely and at term. The American Journal of Clinical Nutrition, 34, 258-265. Anderson, D. (1999). Nutrition for premature infants. In P.Queen Samour, K.King Helm, and E. Lang (Eds.), Handbook of pediatric nutrition (pp43-63). Maryland: Aspen Publications. Anderson, D. (2002). Nutritional assessment and therapeutic interventions for the premature infant. Clinics of Perinatology, 29, 313-326. Atkinson, S.A., Radde, I., Chance, G., Bryan, M., & Anderson, G. (1980). Macro-mineral content of milk obtained during early lactation from mothers of premature infants. Early Human Development, 4, 5-14. Atkinson, S.A., Bryan, M.H., & Anderson, G.H. (1981). Human milk feeding in premature infants: Protein, fat and carbohydrate balances in the first 2 weeks of life. The Journal of Pediatrics, 99, 617-624. Atkinson, S.A. (1994). Nutritional requirements of premature infants. Infant Nutrition for the 90s, 4, 1-3. Anderson, D. (2000). Nutritional implications of premature birth, birth weight, and gestational age classification. In S.Groh-Wargo, M. Thompson, and J.Cox (Eds.), Nutritional care for high-risk newborns (pp 3-10). Chicago: Precept Press. Arbuckle, T.E. , Wilkins, R., & Sherman, G.J. (1993). Birth weight percentiles by gestational age in Canada. Obstetrics and Gynecology, 81, 39-48. Babson, S.G. , & Benda, G.l. (1976). Growth graphs for the clinical assessment of infants of varying gestational age. The Journal of Pediatrics, 89, 814-820. Bach, A.C. , & Babayan, V.K. (1982). Medium-chain trigylcerides: an update. The American Journal of Clinical Nutrition, 36, 950-962. 109 References Brady, M.S., Rickard, K.A., Ernst, J.A., Shreiner, L , & Lemons, J.A. (1982). Formulas and human milk for premature infants: A review and update. Journal of the American Dietetic Association, 81, 547-552. British Columbia Vital Statistics Agency, Ministry of Health and Ministry Responsible for Seniors. (1997). Health status indicators in British Columbia: Birth -related and mortality statistics 1991-1995, 9. Brosius, K.K., Ritter, D.A., & Kenny, T.D. (1984). Postnatal growth curve of the infant with extremely low birth weight who was fed enterally. Pediatrics, 74, 778-782. Buesher,.(1994). Necrotizing enterocolitis. Clinics of Perinatology, 21, 247-262. Bitman, J . , Wood, L , Hamosh, M., Hamosh, P., & Mehta, N.R. (1983). Comparison of the lipid composition of breast milk from mothers of term and preterm infants. The American Journal of Clinical Nutrition, 38, 300-312. Butte, N.F., Garza, C , Johnson, C.A., O'Brian Smith, E., & Nichols, B.L. (1984). Longitudinal changes in milk composition of mothers delivering preterm and term infants. Early Human Development, 9, 153-162. Canadian Pediatric Society, Nutrition Committee. (1995). Nutrient needs and feeding of low-birth weight infants. Canadian Medical Association Journal, 152, 1765-1785. Carlson, S.J. , & Ziegler, E.E. (1998). Nutrient intakes and growth of very low birth weight infants. Journal of Perinatology, 18, 252-258. Carlson, S.J. (1999). Development of a pediatric nutrition outcomes management program. Pediatric Nutrition: Building blocks for life, 22, 7-9. Christie, W.W. (1982). Lipid Analysis. New York: Pergamon Press. Clark, R.M., Ferris, A.M., Fey, M., Brown, P.B., Hundrieser, K.E., & Jensen, R.G. (1982). Changes in the lipid content of human milk from 2-16 weeks postpartum. Journal of Pediatric Gastroenterology and Nutrition, 1, 311 -315. Cobo, E. (1973). Effect of differenct doses of ethanol on the milk-ejection reflex in lactating women. American Journal of Obstetrics and Gynecology, 115,817-821. Cohen, J . (1977). Statistical power analysis for the behavioral sciences. New York: Academic Press. Dancis, J . , O'Connel, J.R., & Holt, L.E. (1948). A grid for recording the weight of premature infants. The Journal of Pediatrics, 33, 570-572. Da Silva, P., Knoppert, D.C., Forret, P., & Angelinie, M. (1999). Domperidone administration to increase breast milk production in lactating mothers of premature newborns: A randomized, double-blind placebo controlled trial. [Abstract]. Pediatric Research, 279A. 110 References DeCarvalho, M., Anderson, D., Giangreco, A.,& Pittard, W.B. (1985). Frequency of milk expression and milk production by mothers of nonnursing premature infants. American Journal Diseases in Children, 739,483-485. Dorea, J .G. , Horner, M.R., Bezerra, V.L.V., & Compante, M.L. (1982). Variations in major constituents of foremilk and hindmilk of brazilian women. Journal of Tropical Pediatrics, 28, 303-305. Ehrenkranz, R.A., Ackerman, B.A., & Nelli, CM.(1984). Total lipid content and fatty acid composition of preterm human milk. Journal of Pediatric Gastroenterology and Nutrition, 3, 755-758. Ehrenkranz, R.A., & Ackerman, B. (1986). Metoclorpramide effect of faltering milk production by mothers of premature infants. Pediatrics, 78, 614-620. Ehrenkranz, R.A., Younes, N., Lemons, J.A., Fanaroff, A.A., Donavon, E.F., Wright, L.L., et al. (1999). Longitudinal growth of hospitalized very low birth weight infants. Pediatrics, 104, 280-289. Ehrenkranz, R.A. (2002). Growth outcomes of very low-birth weight infants in the newborn intensive care unit. Clinics of Perinatology, 27, 325-344. El-Mohandes, A.E . , Picard, M.B., & Simmens, S. (1997). Use of human milk in the intensive care nursery decreases the incidence of nosocomial sepsis. Journal of Perinatology 17, 130-134. Emery, W.B., Canolty, N.L., Aitchinson, J.M., & Dunkley, W.C. (1978). Influence of sampling on fatty acid composition of human milk. The American Journal of Clinical Nutrition, 31, 1127-1130. European Society of Pediatric Gastroenterology and Nutrition. (1987). Nutrition and feeding of preterm infants. Acta Paediactrica Scandinavica, 336 (Suppl.), 3-14. Evans, R.A., & Thureen, P.J. (2001). Early feeding strategies in preterm and critically ill neonates. Neonatal Network, 20, 7-18. Feher, S., Berger, L., Johnson, J.,& Wilde, J . (1989). Increasing breast milk production for premature infants with a relaxation/imagery audiotape. Pediatrics, 83, 57-60. Field, A. (2000). Discovering Statistics using spss for windows. London: Sage Publications. Folch, J . , Lees.M., & Stanley, S. (1957). A simple method for the isolation and purification of total lipids from animal tissues. Journal of Biological Chemistry, 226, 497-509. Fomon, S. (1993). Nutrition of normal infants. Baltimore: Mosby-Year Book. 111 References Forsum, E., & Lonnerdal, B. (1979). Variation in the contents of nutrients of breast mlk during one feeding. Nutrition Reports International, 19, 815-820. Fulhan, J.K. (2000). Prematurity. In K.M. Hendricks, C. Duggan, & W.A. Walker (Eds.), Manual of pediatric nutrition (pp 502-520). London: B.C. Decker. Gibbs, J.H., Fisher, C , Bhattachyarya, S., Goddard, P., & Baum, J.D. (1977). Drip breast milk: its composition, collection and pasteurization. Early Human Development, 1, 227-245. Green, D., Moye, L., Schreniener, R.L., & Lemons, J.A.(1982). The relative efficacy of four methods of human milk expression. Early Human Development, 6,153-159. Griffin, T.L., Bradford, L.P., & Engstrom, J.L. (2000). Mothers' performing creamatocrit measures in the nicu: Accuracy, reactions and cost. Journal of Obstetric, Gynecologic and Neonatal Nursing, 29, 249-257. Gross, S.J. , David, R.J., Bawman, L., &Tomarelli, R.M.(1980). Nutritional composition of milk produced by mothers delivering premature. The Journal of Pediatrics, 96, 641-644. Groh-Wargo, S., Toth, A., Mahoney, D., Simonian, S., Wasser, T., & Rose, S. (1995). The utility of a bilateral breast pumping system of mothers of premature infants. Neonatal Network, 74,31-36. Groh-Wargo, S..Thompson, M.,Cox, J.(2000). Nutritional care for high-risk newborns. Chicago: Precept Press. Hall, B. (1979). Uniformity of human milk. The American Journal of Clinical Nutrition, 32, 304-312. Hamosh, M. (1994). Digestion in the premature infant: The effects of human milk. Seminars of Perinatology, 18, 485-494. Hassard, T.M. (1991). Understanding biostatistics. Baltimore: Mosby-Year Book. Heird, W.C., Driscoll, J.M., Schullinger, J.N., Grebin, B., & Winters, R.W. (1972). Intravenous alimentation in pediatric patients. Journal of Pediatrics, 80, 351-372. Hill, P., Aldag, J . , & Chatterton, R. (1996). The effect of sequential and simulateous breast pumping on milk volume and prolactin levels:A pilot study. Journal of Human Lactation, 12, 193-199. Hill, P., Aldag, J . , & Chatterton, R. (1999). Effects on pumping style on milk production in mothers of nonnursing preterm infants. Journal of Human Lactation, 15, 209-216. Hopkinson, J . , Schanler, R., & Garza, C. (1988). Milk production by mothers of premature infants. Pediatrics, 81, 815-820. 112 References Hopkinson, J . , Schanler, R., Fraley, K., & Garza, C. (1992). Milk production by mothrs of premature infants: Influence of cigarette smoking. Pediatrics, 90, 934-938. Hovasi Cox, J . (2000). Bronchopulmonary dysplasia. In S.Groh-Wargo, M. Thompson, and J.Cox (Eds.), Nutritional care for high-risk newborns (pp 369-390). Chicago: Precept Press. Howard, C , & Lawrence, R. (1999). Drugs and breastfeeding. Clinics in Perinatology, 26, 447-478. Howell, D.C. (1997). Statistical Methods for Psychology. Belmont ,CA: Wadsworth Publishing. Hurst, N., Valentine, C , Renfro, L., Burns, P.,& Ferlic, L. (1997). Skin to skin holding in the neonatal intensive care unit influences maternal milk volume. Journal of Perinatology, 17, 213-217. Hurst, N. (1999). Expression and handling of human milk. In Report of the 108 th Ross Conference on Pediatric Research. Human milk for very low birth weight infants (pp 120-128). Columbus, OH: Ross Products Division, Abbott Laboratories. Hytten, F.E. (1954a). Clinical and chemical studies in human lactation. I.Collection of milk samples. British Medical Journal, 1, 175-176. Hytten, F.E. (1954b). Clinical and chemical studies in human lactation. II. Variation in major constituents during a feeding. British Medical Journal, 1, 176-179. Hytten, F.E. (1954c). Clinical and chemical studies in human lactation. III.Diurnal variations in major constituents in milk. British Medical Journal, 1, 179-182. Hytten, F.E. (1954d). Clinical and chemical studies in human lactation. IV.Trends in milk composition during the course of lactation. British Medical Journal, 1, 249-253. Hytten, F.E. (1954e). Clinical and chemical studies in human lactation. V. Individual differences in composition of milk. British Medical Journal, 1, 253-255. Jensen, R.G. (1989a). The lipids of human milk. Boca Raton,FL: C R C Press. Jensen, R.G.(1989b). Lipids in human milk - composition and fat-soluble vitamins. In E. Lebenthal (Ed.), Textbook of Gastroenterology and Nutrition in Infancy (pp 157-208). New York:Raven Press. Jensen, R.G. (1999). Lipids in Human Milk. Lipids, 34, 1243-1271. Jones Wessel, J . (2000). Feeding methodologies. In S.Groh-Wargo, M. Thompson, and J.Cox (Eds.), Nutritional care for high-risk newborns (pp 321-339 ). Chicago: Precept Press. 113 References Katrine, K.F. (2000). Anthropometric assessment. In S. Groh-Wargo, M. Thompson, and J.Cox (Eds.), Nutritional care for high-risk newborns (pp 11-12). Chicago: Precept Press. Kavanaugh, K., Meirer, P., Zimmerman, B.(1997). The rewards outweigh the efforts breastfeeding outcomes for mothers of preterm infants. Journal of Human Lactation, 73,15-21. Khachatryan, A., & Linardakis, N. (2000). Biostatistics and Epidemiology. New York: McGraw-Hill. Krause, M.V., & Mahan, L.K. (1984). Food, nutrition, and diet therapy. Philadelphia: W.B. Saunders Company. Kristen, D., & Bradford, L. (1999). Hindmilk feedings. Neonatal Network, 18, 68-70. Kubit, J .G. (2000). Lactation issues. In S.Groh-Wargo, M. Thompson, and J.Cox (Eds.), Nutritional care for high-risk newborns (pp303-319). Chicago: Precept Press. Lammi-Keefe, C.J . , & Jensen, R.G. (1984). Lipids in human milk: A review. 2: Composition and fat-soluble vitamins. Journal of Pediatric Gastroenterology, 3, 172-198. Lang, S. (1997). Breastfeeding special care babies. London: Bailliere Tindall. Lau, C. (2001). Effects of stress on lactation. Pediatric Clinics of North America,48, 221-234. Lawrence, R.A, % Lawrence, R.M.(1999) Breastfeeding: A guide for the medical profession. Baltimore: Mosby. Lemons, J.A., Schreiner, R.L., & Gresham, E.F. (1980). Simple method for determining the caloric and fat content of human milk. Pediatrics, 66, 626-628. Lemons, J.A., Moye, L., Hall, D., & Simmons, M. (1982). Differences in the composition of preterm and term human milk during early lactation. Pediatric Research, 16,113-117. LePage, G., Collet, S., Bougie, D., Kien, L.C., & Roy, C C . (1984). The composition of preterm milk in relation to the degree of prematurity. The American Journal of Clinical Nutrition, 40, 1042-1049. LePage, G., & Roy, C C (1988). Specific methylation of plasma nonesterified fatty acids in a one-step reaction. Journal of Lipid Research, 29, 277-235. Lubchenco, L.O., Hansman, C , Dressier, M., & Boyd, E. (1963). Intrauterine growth as estimated from liveborn birth-weight data at 24-42 weeks of gestation. Pediatrics, 32, 793-800. 114 References Lucas, A., Gibbs, J.A.H., Lyster, R.L., & Baum, J.D. (1978). Creamatocrit: Simple clinical technique for estimating fat concentration and energy value of human milk. British Medical Journal, 22,1018-1020. Lucas, A., & Cole, T.J. (1990). Breast milk and neonatal necrotising enterocolitis. The Lancet,336, 1519-1523. Lucas, A., Morley, R., Cole, T.J. , Lister, G., & Leeson-Payne, C. (1992). Breast milk and subsequent intelligence quotient in children born preterm. The Lancet 339, 261-264. Macy, I.G., Nims, B., Brown, M., & Hunscher, H.A. (1931). Human milk studies Vll. Chemical analysis of milk representative of the entire first and last halves of the nursing period. American Journal of Diseases of Children, 42, 568-589. Malloy, M.H., & Graubard, B. (1993). Predictors of rehospitalization among very low birth weight infants.[Abstract]. Clinical Research 41, 791 A. Meier, P., & Brown, L. (1996). State of the science: Breastfeeding for mothers and low birth weight infants. Nursing Clinics of North America 3, 351-365. Meier, P. (1998). Extremely low birthweight infants: Individualizing lipid content of own mothers' milk. ABM News and Views, 4, 5-6. Meier.P., Brown, L., & Hurst, N. (1999). Breastfeeding the preterm infant. In Riordan, J . and Auerbach,K. (Eds.). Breastfeeding and Human Lactation (pp. 449-480).Toronto: Bartlett Publishers. Meier, P.P., Murtaugh, M., Vasan, U., Meier, W. & Schanler, R.J. (1999). Modification of the lipid concentration in own mothers' milk (OMM) feedings in the neonatal intensive care unit (NICU): Clinical application of the creamatocrit (CRT) technique. [Abstract]. Pediatric Research, 45, 287A. Micheli, J . & Schutz, Y. (1993). Protein. In R.C. Tsang, A. Lucas, R. Uauy, and S. Zlotkin (Eds.), Nutritional Needs of the Preterm Infant: Scientific Basis and Practical Guidelines (pp 22-46). Baltimore: Cadueceus Medical Publishers. Munro, B. Visintainer, M.A., & Page E. (1986). Statistical methods for health care research. Philadelphia: J.B. Lippincott. Narayanan, I., Prakash, K., & Gujral, V.V. (1981). The value of human milk in the prevention of infection in the high-risk low birth weight infant. Journal of Pediatrics 99, 496-498. Narayanan, I., Brakash, K., & Prabhakar, A.K. (1982). A planned prospective evaluation of the anti-infective property of varying quantities of expressed human milk. Acta Paediatrica Scandinavica, 71, 441-445. 115 References Nelville, M.C., Keller, R.P., Seacat, J . , Casey, C .E . , Allen, J . C , & Archer, P. (1984). Studies on human lactation. I. Within-feed and between breast variation in selected components of human milk. The American Journal of Clinical Nutrition, 40, 635-646. Nevin-Folino, N.L., Loughead, J.L., & Loughead, M.K. (2001). Enhanced -calorie formulas: Considerations and options. Neonatal Network, 20, 7-15. Newkirk, M., Fulton, J . , & Schmidt, K. (1999). Nutrition management of bronchopulmonary dysplasia, BPD. Building Blocks for Life: Pediatric Nutrition, 22, 1-4. Norman, G.R., & Streiner, D.L. (1994). Biostatistics: The bare essentials. Baltimore: Mostby-Year Book. Noss Whitney, E., & Rody Rolfes, S. (1999). Understanding Nutrition. USA: Wadsworth Publishing. Ntoumanis, N. (2001). A step-by-step guide to spss for sport and exercise studies. New York: Routledge. Pauls, J . , Bauer, K., & Wersmold, H. (1998). Postnatal body weight curves for infants below 1000g birth weight receiving early enteral and parenteral nutrition. European Journal of Pediatrics, 157, 416-421. Pereira, G.R. (1995). Nutritional care of the extremely premature infant. Clinics in Perinatology, 22, 61-75. Prentice, A., Prentice, A.M., & Whitehead, R.G. (1981). Breast-milk fat concentrations of rural African women. 1 .Short-term variations within individuals. British Journal of Nutrition, 45, 483-494. Raiha, N.C.R. (1999). Individualized protein fortification of human milk. In Report of the 108 th Ross Conference on Pediatric Research. Human milk for very low birth weight infants (pp 144-145). Columbus, OH: Ross Products Division, Abbott Laboratories. Rose, J . , Gibbons, K., Carlson, S.E. , & Koo, W.W. K. (1993). Nutrient needs of the preterm infant. Nutrition in Clinical Practice, 8, 226-232. Salkind, N.J. (2000). Statistics for people who think they hate statistics. Thousand Oaks, CA:Sage Publications. Sapsford, A.L. (2000). Human milk and enteral nutrition products. In S.Groh-Wargo, M. Thompson, J . Cox (Eds.), Nutritional care for high-risk newborns (pp265-302). Chicago:Precept Press. Schanler, R., Oh, W. (1980). Composition of breast milk obtained from mothers of premature infants as compared to breast milk obtained from donors. Pediatrics, 96, 679-681. 116 References Schanler, R. (1995). Suitability of human milk for the low-birthweight infant. Clinics in Perinatology, 22, 207-222. Schanler, R., Hurst, N., & Lau, C. (1999). The use of human milk and breastfeeding in premature infants. Clinics in Perinatology, 26, 379-398. Schanler, R. (2001). The use of human milk for premature infants. Pediatric Clinics of North America, 48,207-219. Slusher, T., Hampton, R., Ishaya, A., Bode-Thomas, F., Pam, S., Akor, F., et al. (2000). Hindmilk feedings for hospitalized low birthweight infants in nigeria: Effect on infant weight gain and maternal milk volume. [Abstract].Pedaitric Research, 47, 297A. Simmer, K., Metcalf, R., & Daniels, L.(1997). The use of breastmilk in a neonatal unit and its relationship to protein and energy intake and growth. Journal of Pediatric Child Health ,33, 55-60. Sparks, J.W. (1984). Human intrauterine growth and nutrient accretions. Seminars in Perinatology, 8, 74-93. Spicer, K. (2001). What every nurse needs to know about breast pumping:lnstructing and supporting mothers of premature infants in the nicu. Neonatal Network, 20, 35-41. Tsang, R.C., Lucas, A., Uauy, R., & Zlotkin, S.(Eds.)(1993). Nutritional Needs of the Preterm Infant. Baltimore: Caduceus Medical Publishers. Uraizee, F., & Gross, S.J. (1989). Improved feeding tolerance and reduced incidence of sepsis in sick very low birthweight (vlbw) infants fed maternal milk. [Abstract]. Pediatric Research 25, 298A. Valentine, C.J. , Hurst, N.M., & Schanler, R.J. (1994). Hindmilk improves weight gain in low-birth-weight infants fed human milk. Journal of Pediatric Gastroenterolgy and Nutrition, 18, 474-477. Vasan, U., Meier, P., Meier, W., & Kristen, D. (1998). Individualizing the lipid ontent of own mothers' milk: Effect on weight gain for extremely low birth weight (elbw) infants.[Abstract]. Pediatric Research, 43, 270A. Wang, C D . , Chu, P.S., Meller, B.G, & Shenai, J.P. (1999). Creamatocrit and the nutrient composition of human milk. Journal of Perinatology, 19, 343-346. Watt, B.K., & Merril, A.L. (1964). Composition of foods: Raw, processed, prepared. Agriculture Handbook No.8. Washington: United States Department of Agriculture. Winters, C.H., Hoving, E.B., & Muskiet, A.J . (1993). Fatty acid composition of human milk triglyceride species. Journal of Chromatography, 616, 9-24. Wilson, D. (1995). Nutrition of the preterm infant. British Journal of Obstetrics and Gynaecology, 102, 854-860. 117 References Wright, K., Dawson, J.P., Fallis, D., Wogt, E., & Lorch, V. (1993). New postnatal growth grids for very low birth weight infants. Pediatrics, 91, 922-926. Ziegler, E.E., Thureen, P.J., & Carlson, S.J. (2002). Aggressive nutrition of the very low birthweight infant. Clinics of Perinatology, 29, 225-244. Zolman, J.F. (1993). Biostatistics:Experimental design and statistical inference. New York: Oxford University Press. 118 Appendix 1 APPENDIX 1 Pump Room Recruitment Letters 119 Appendix 2 APPENDIX 2 Bedside Recruitment Letter 123 Appendix 3 APPENDIX 3 Study Consent Forms 127 Appendix 3 I N F O R M E D C O N S E N T Development of a Protocol for Fractionating Expressed Human Milk into Foremilk and Hindmilk Principle Investigator: Dr. She i l a M. Innis Co-investigator: Dr. John Smyth Summary of Project: Fat is the major source of calor ies in human milk, representing 40 -50% of the total calor ies. Fat is the most variable component of human milk. It var ies throughout the day, within a feed (from foremilk to hindmilk), with stage of lactation, and from woman to woman . The fat content of foremilk, milk at the beginning of a feed, is lower than at the end of the feed (hindmilk). For some infants who grow poorly due i l lness, high energy requirements, fluid restrictions, or prematurity, having milk which is higher in fat content, and hence calor ies, may be beneficial for promoting weight gain. Purpose of Study: The purpose of this study is to develop and evaluate guidel ines for separat ing mother's milk into 2 fractions: foremilk and hindmilk. Three different methods, based on visual appearance , the milk express ion time and the milk vo lume expressed will be compared for practicality and fat content. In order to participate in this study you need to be express ing at least 120 mL daily of mature breastmilk. Study Procedure: This study involves no risk. I understand that by participating, I will: 1. Keep a 24 hour milk express ion diary for the basel ine and for each of the three method. 2. Fol low usual S C N guidel ines for set-up, handling and storage of expressed milk. 3. Take a smal l amount of my expressed milk during the basel ine period (1-2 mL from each container pumped during the 24 hour period) and give the samples to the nutrition researcher for analysis of fat content. 4. Separate my expressed milk (during each pumping session) into foremilk and hindmilk using three different methods. E a c h method requires a 24 hour collection period. To help learn the procedures for separat ing milk, a 24 hour "practice day" is needed prior to the 24 hour col lection period. 5. Take a smal l amount of my expressed milk (1-2 mL sample of foremilk and hindmilk from each container pumped during the 24 hour period) and give the samples to the nutrition researcher for the analys is of fat content. 6. At my discretion complete a confidential survey on soc io-demographic information. Page 1 of 2 128 Append ix 3 I N F O R M E D C O N S E N T Validation of a Protocol for Fractionating Expressed Human Milk into Foremilk and Hindmilk Principle Investigator: Dr. She i l a M. Innis Co-investigator: Dr. John Smyth Summary of Project: Fat is the major source of calor ies in human milk, representing 40 -50% of the total calor ies. Fat is the most variable component of human milk. It var ies throughout the day, within a feed (from foremilk to hindmilk), with stage of lactation, and from woman to woman . The fat content of foremilk, milk at the beginning of a feed, is lower than at the end of the feed (hindmilk). For some infants who grow poorly due i l lness, high energy requirements, fluid restrictions, or prematurity, having milk which is higher in fat content, and hence calor ies, may be beneficial for promoting weight gain. Purpose of Study: The purpose of this study is to validate a new protocol deve loped for separat ing mother's milk into 2 fractions: foremilk and hindmilk. In order to participate in this study you need to be express ing at least 120 mL daily of mature breastmilk. Study Procedure: This study involves no risk. I understand that by participating, I will: 1. Keep a 24 hour milk express ion diary for the basel ine and for the study period using a new protocol deve loped for separat ing expressed milk into foremilk and hindmilk. 2. Fol low usual S C N guidel ines for set-up, handling and storage of expressed milk. 3. Take a smal l amount of my expressed milk during the basel ine period (1-2 mL from each container pumped during the 24 hour period) and give the samples to the nutrition researcher for analysis of fat content. 4. Separate my expressed milk (during each pumping sess ion within the 24 hour period) into foremilk and hindmilk for 3 nonconsecut ive days (i.e. 2 weekdays and 1 weekend day) using the new fractionating protocol. To help learn the procedures for separat ing milk, a 24 hour "practice day" is needed prior to the 24 hour collection period. 5. Take a smal l amount of my expressed milk (1-2 mL sample of foremilk and hindmilk from each container pumped during the 24 hour period) and give the samples to the nutrition researcher for the analysis of fat content. 6. At my discretion complete a confidential survey on soc io-demographic information. P a g e 1 of 2 130 Appendix 3 INFORMED C O N S E N T Determination of Expressed Breastmilk Practices of Mothers Who Deliver Premature Infants Principle Investigator: Dr. Sheila M. Innis Co - Investigator: Dr. John Smyth Summary of Project: Breastmilk is the optimal choice of feeding for infants because of its many nutritive and non-nutritive components. Many premature infants receive expressed breastmilk until they are able to breastfeed. Mothers of premature infants may express their breastmilk for many weeks to months. There can be large variations between mothers in the volume of milk produced. Several factors can influence a mother's breastmilk supply. The frequency and duration of pumping, stress and other factors influence the volume of breastmilk expressed daily. Purpose of Study: The purpose of this study is to look at breastmilk expression practices and factors that may affect breastmilk production of mothers who deliver premature infants. This information will be used for nutrition studies that involve the expression of breastmilk for premature infants, as well as supporting mothers in maintaining optimal milk supply. In order to participate in this study, you need to be expressing mature (> 14 days postpartum) breastmilk. Study Procedure: The study involves no risk. I understand that by participating, I will: 1) Keep a 3 day recording diary to collect information on breastmilk expression practices for 3 non-consecutive days (2 weekdays and 1 weekend day). 2) Follow usual S C N guidelines for set-up, handling, storage, and transport of expressed breastmilk. 3) Provide information on my usual pumping routine and breastmilk supply. 4) Provide information on possible factors that may affect my pumping routine and breastmilk supply. 5) At my discretion complete a confidential survey on socio-demographic information. Page 1 of 2 132 Appendix 4 APPENDIX 4 Phase I Nutrition Study Booklet 134 Appendix 4 INSTRUCTIONS FOR SEPARATING, COLLECTING AND LABELING EXPRESSED BREASTMILK 1) Follow the SCN guidelines for set-up, handling and storage of expressed milk. 2) FRACTIONATING INSTRUCTIONS: follow the instructions provided for each method on how to separate milk into foremilk and hindmilk. 3) Label the containers as FOREMILK and HINDMILK. If more than one container is needed to collect Foremilk or Hindmilk, indicate the number of containers pumped in order of expression (i.e. Hindmilk #1/2, Hindmilk #2/2). If double pumping, you will have foremilk and hindmilk containers from each breast. 4) Using a syringe, take 1-2 mL sample of milk (from each container) of foremilk and hindmilk expressed during the 24 hr study period. These samples will be analyzed for fat content. (Note: if the sample is not taken immediately after pumping, shake the container gently to mix the fat before drawing milk up into the syringe) 5) Label the samples (syringes) as FOREMILK and HINDMILK with the labels provided. Fill in the following information: (Note: color coding system: blue = foremilk, yellow = hindmilk) DATE: TIME: NAME: # / F O R E / H I N D MTHD: 1 f Circle List the # Gradufeeds pumped (i.e. #1/2, #2/2) List the Method #: #1= Visual method #2= Time method #3= Volume method 6) Store samples in the refrigerator on the unit for the nutrition researcher to collect. 7) If pumping at home, refrigerate samples until you bring them to the hospital. Keep the samples cold for transporting by using a small "cooler chest" or thermal lunch bag with an icepack inside. 8) Once foremilk and hindmilk samples have been collected for the study, you can combine the milk into one container for storage and/or feeding. 136 BASELINE INFORMATION Appendix 4 A 24 hour milk expression diary is needed to collect baseline information on your usual pumping routine and volume of milk expressed. A sample of expressed milk will be collected during a 24 hour baseline period to analyze the fat content of your milk. * Use the table provided to record the baseline information. * Follow the " Instructions for Collecting, Storing and Labeling Expressed Breastmilk" on page 2. MILK EXPRESSION DIARY (Baseline) Date: TIME of Day DURATION of pumping session (minutes) TYPE OF PUMP (electric/ hand) METHOD of pumping (single/double) SIDE of Breast (right=R, left=L, both=B) LOCATION of pumping session (home, hospital, other) VOLUME of milk expressed (per container per breast) (mL) TOTAL VOLUME of milk expressed (mL) Example: 8:00 am 20 mins electric double both hospital R= 30 mL L = 60mL 90 mL 137 Appendix 4 Please answer the following questions regarding breastmilk expression practices and possible factors that affect breastmilk supply. 1) List any factors that you feel may affect your milk expression routine. Eg. A busy home schedule makes it difficult for me to express regularly. 2) List any factors that you feel may affect your milk volume supply and indicate whether these factors increase or decrease your milk volume supply. Eg. I think pumping every 4 hours increases my milk supply. 3) List any factors that you feel may affect your "letdown" reflex (increased milk flow). Eg. Seeing my baby starts my letdown immediately. 4) From the list below, indicate if you do anything special to help stimulate your milk production/secretions. (You may check more than one box.) massage breast touching, seeing, hearing baby relaxation techniques pumping regularly reposition pump on breast < no special techniques used thinking of your baby/picture of baby other: 5a) Does your expressed milk volume change from day to day? yes no -> go to #6 don't know -> go to #6 a) If yes, what is your range of milk volume expressed daily? Eg. 300-400 mL /day mL/day 138 Appendix 4 "Letdown" is an increased flow of milk that may occur during or before a pumping session? The following questions (#6 a,b,c) are about the "letdown" reflex. 6) Do you experience "letdown"? yes no —> go to #7 don't know-* go to #7 If you experience "letdown", please answer the following: a) When do you experience "letdown"? (if you experience more than one "letdown" during a pumping session, indicate by checking more than one box) before session beginning session middle session end session other: b) How often do you experience "letdown" when pumping? always ( all the time) most of the time ( more than half the time) sometimes (approx. half the time) occasionally (less than half the time) other: c) The following is a list of some signs/signals that "letdown" has occurred. Please indicate by checking which box best describes how you tell when "letdown " has occurred, (you may check more than one box) tingling sensation warm feeling "gushing" feeling increased volume other: 7) What is your daily fluid intake? (1 cup=8 oz) cups per day 8) Are you presently taking any prescribed medications that increase your breastmilk supply? yes no -> go to #9 don't know -> go to #9 If yes, please list the name of the medication: 139 Appendix 4 9) Are you presently taking any other products that specifically increase your breastmilk supply? Eg. herbs or herbal products yes no -> go to #10 don't know -> go to #10 If yes, please list the name of the product: 10) Are you presently taking any other prescription medications (other than those listed above) or any non-prescription medications? yes no -> go to #11 don't know-» go to #11 If yes, please list: The following questions are about drinking caffeinated beverages (i.e. coffee, tea, cola-drinks) and herbal beverages. If you do not drink these beverages go to question #13. 11) In the past 4 weeks, have you consumed any caffeinated beverages? yes no —> go to #12 don't know —> go to #12 If yes, please answer the following: a) What type(s) of beverage did you consume? (check all that apply) coffee tea cola-type drink other: b) What is the average number of cups (1 cup=8 oz) you drink per week? cups 12) In the last 4 weeks, have you consumed any herbal beverages (i.e. herbal teas) or specialty teas? yes no -> go to #13 don't know go to #13 If yes, please answer the following: a) List the types of herbal beverage or specialty tea. b) What is the average number of cups (1 cup=8 oz) you drink per week? cups 140 Appendix 4 The following questions are about alcohol use. If you do not drink alcohol please go to question #14. The term "drink" refers to 12 oz beer, 5 oz wine or 1.5 oz spirit/hard liquor. 13) In the past 4 weeks, have you taken a drink of alcohol? yes no - » go to #14 don't know -> go to #14 If yes, please answer the following: a) What is the average number of drinks consumed? per month b) What type of alcoholic beverage? (check all that apply) beer/cider/cooler wine spirits/hard liquor other: The following questions are about cigarette smoking. If you do not smoke cigarettes, go to question #15. 14a) At the present time, do you smoke cigarettes? yes no b) What is the average number of cigarettes you smoke per week? Many things can contribute to feelings of stress, including having a baby in the hospital. 15) Are you presently experiencing any stress? yes -» mild stress no -» go to #16 don't know -» go to #16 moderate stress severe stress b) If you are experiencing stress, do you feel this stress is affecting your i) milk supply yes no don't know ii) "letdown" reflex yes no don't know iii) milk expression routine yes no don't know 16) Please provide any additional comments on your breastmilk expression practices or factors you feel affect your breastmilk supply? 141 VISUAL METHOD (Method #1) Appendix 4 A 24 hour diary and sample of milk are needed to collect information on how to separate your expressed breastmilk into 2 fractions: foremilk (milk expressed at the beginning of a feed) and hindmilk (milk expressed for the remainder of the feed) based on the VISUAL METHOD. A small sample of foremilk and hindmilk will be collected to analyze the fat content. The VISUAL METHOD involves determining when your breastmilk changes in color and consistency from a thin, white milk to a thick, creamier milk. The first fraction of milk expressed will be the thin, white milk and will be called "foremilk". The remaining milk (thick, creamier milk) expressed will be called "hindmilk". INSTRUCTIONS: 1) Follow "Instructions for Separating, Collecting and Labeling Expressed Breastmilk" on page 2. 2) Use the VISUAL METHOD to separate expressed milk into foremilk and hindmilk. The "point of separation" will be when you see a change in the milk color and consistency from a thin, white milk (foremilk) to a thick, creamier milk (hindmilk). You need a watch or clock close by to note the time. © 3) Change the container (gradufeed) at the "point of separation". The first fraction of milk expressed will be called FOREMILK. The remaining milk expressed will be called HINDMILK. 4) Fill out the table on the next page for the Visual Method. When recording the volume of milk expressed at each session, list which breast is pumped (R=right; L=left) and list the amount of milk collected from each breast. If more than one container is used to collect milk from a breast, indicate the order of expression by using numbers. For example: R1 (right breast, 1st container) and R2 (right breast 2nd container) COMMENTS: 142 Appendix 4 Order: 1 s t - 2 n d - 3 r d VISUAL METHOD (Method #1) Date: TIME of Day DURATION of pumping session (minutes) TYPE of pump (hand or electric) METHOD Of pumping (single/d ouble) SIDE of breast (right=R, left=L, both=B) LOCATION of pumping session (home, hospital, other) LENGTH OF TIME to change in color, consistency (minutes) VOLUME of milk expressed (list for each container) (R= right breast; L = left breast) Fore (ml) Hind(mL) E x a m p l e 8 :00 a m 2 0 m i n s E l e c t r i c D o u b l e B o t h H o s p i t a l 3 m i n s R = 2 0 L = 3 0 R i = 6 0 R 2 = 1 0 L 1 = 5 0 Did you experience "letdown" during the pumping session? yes no don't know If you experienced "letdown", did it occur before or after the change in color, consistency of the milk? before after other: 143 TIME (Letdown) METHOD (Method # 2 ) Appendix 4 A 24 hour diary and sample of milk are needed to collect information on how to separate your expressed breastmilk into 2 fractions: foremilk (milk expressed at the beginning of a feed) and hindmilk (milk expressed for the remainder of the feed) based on the TIME (Letdown) METHOD. A small sample of foremilk and hindmilk will be collected to analyze the fat content. The TIME (Letdown) METHOD involves determining when "letdown" (increased milk flow) has occurred, after you start pumping. You will continue to pump for 2-3 minutes after "letdown". The milk expressed in this first fraction will be called "foremilk". Milk expressed after this point will be called "hindmilk". INSTRUCTIONS: 1) Follow "Instructions for Separating, Collecting and Labeling Expressed Breastmilk" on page 2. 2) Use the TIME (Letdown) METHOD to separate expressed milk into foremilk and hindmilk. The "point of separation" will be 2-3 minutes after "letdown" occurs. (Note: the exact time will be confirmed with the researcher) You need a watch or clock close by to note the time. © If you do not experience "letdown", the point of separation will be 2-3 minutes after the start of pumping. 3) Change the container (gradufeed) at the "point of separation". The first fraction of milk expressed will be called FOREMILK. The remaining milk expressed will be called HINDMILK. 4) Fill out the table on the next page for the Time Method. When recording the volume of milk expressed at each session, list which breast is pumped (R=right; L=left) and list the amount of milk collected from each breast. If more than one container is used to collect milk from a breast, indicate the order of expression by using numbers. For example: R1 (right breast, 1st container) and R2 (right breast 2nd container) COMMENTS: 144 Appendix 4 Order: 1s t - 2 n d - 3 r d TIME METHOD (Method #2) Date: Expressed Breastmilk will be separated at: minutes (after letdown/after start pumping) TIME of Day DURATION of pumping session (minutes) TYPE of pump (hand or electric) METHOD Of pumping (single/ double) SIDE of breast (right=R, left=L, both=B) LOCATION of pumping session (home, hospital, other) LENGTH OF TIME to "Letdown" (minutes) VOLUME of milk expressed (list for each container) (R= right breast; L= left breast) Fore I Hind Example 8:00 am 20 mins Electric Double Both Hospital 2 mins R=20 L=30 R1=60 R2=10 L1=50 145 VOLUME METHOD (Method #3) Appendix 4 A 24 hour diary and sample of milk are needed to collect information on how to separate your expressed breastmilk into 2 fractions: foremilk (milk expressed at the beginning of a feed) and hindmilk (milk expressed for the remainder of the feed) based on the VOLUME METHOD. The VOLUME METHOD involves separating milk based on a set volume or proportion of milk expressed. The total volume of milk you express daily and your baby's daily fluid intake will be used to calculate the volume or proportion of milk to be called "foremilk". The remaining milk expressed will be called "hindmilk". Your previous day's expressed milk volume will be a guide as to how much milk is typically obtained during a session. This information will help to determine how much volume will be "foremilk". % value = Mom's daily milk volume expressed - your baby's daily milk intake x 100% Mom's daily milk volume expressed For example: i) Mom's daily milk volume expressed: 600 mL ii) Your baby's daily milk intake: 360 mL iii) Difference: expressed - intake 240 mL (extra milk expressed daily) iv) % value (240 + 600 mL x 100%) 40 % v) The first 40% of milk expressed at each session will be called "foremilk". vi) Round off (up or down) volumes of milk to the nearest 5 mL SESSION # PREVIOUS DAY VOLUME EXPRESSED 40% VOLUME FOREMILK (1st PORTION) (rounded off values) 1 120 mL 48 mL 50 mL 2 100 mL 40 mL 40 mL 3 85 mL 35 mL 35 mL 4 100 mL 40 mL 40 mL 5 105 mL 42 mL 40 mL 6 90 mL 36 mL 35 mL Total = 600mL Total = 240 mL INSTRUCTIONS: 1) Follow "Instructions for Separating, Collecting and Labeling Expressed Breastmilk" on page 2. 2) Use the VOLUME METHOD to separate expressed milk into foremilk and hindmilk. The "point of separation" will be based on a proportion or set volume of milk expressed at the beginning of a pumping session (foremilk). You may need a calculator to help you. 146 Appendix 4 3) Change the container (gradufeed) at the "point of separation". The first fraction of milk expressed will be called FOREMILK. The remaining milk you express will be called HINDMILK. 4) Fill out the table on the next page for the Volume Method. When recording the volume of milk expressed at each session, list which breast is pumped (R=right; L=left) and list the amount of milk collected from each breast. If more than one container is used to collect milk from a breast, indicate the order of expression by using numbers. For example: R1 (right breast, 1st container) and R2 (right breast 2nd container) NOTE: It may be helpful to draw a line on the foremilk container (gradufeed) to show the volume designated for the "point of separation". WORKSHEET FOR THE VOLUME METHOD i) Mom's daily milk volume expressed: ii) Your baby's daily milk intake: iii) Difference: expressed - intake iv) % value % value = expressed - intake x 100% expressed v) FOREMILK =The first % of milk expressed at each session SESSION # PREVIOUS DAY VOLUME E X P R E S S E D % VOLUME FOREMILK (1 s t PORTION) (rounded off values) 1 2 3 4 5 6 7 8 Total= mL Total= mL COMMENTS: mL mL mL (extra milk expressed daily) /o 147 Appendix 4 Order: 1s t - 2 n d - 3 r d VOLUME METHOD (Method #3) Date: TIME of Day AMOUNT pumped yesterday at session (mL) VOLUME to be called "foremilk" based on % value (mL) DURATION of pumping session (minutes) TYPE of pump (hand, electric) METHOD of pumping (single, double) SIDE of Breast (right, left, both) LOCATION of pumping session (home, hospital, other) VOLUME of milk expressed (list for each container) (R= right breast; L= left breast) (mL) Fore I Hind xample: 8:00 am 120 mL (R=60) (U60) 50 mL (120x40%) 20 min Electric Double Both Hospital R=25 L=25 R1=35 L1=35 148 Appendix 4 COMMENTS / NOTES: 149 EVALUATION OF THREE FRACTIONATING METHODS Appendix 4 Your feedback is important in the development of a fractionating protocol for our nursery. Please evaluate the three methods you have used to separate your expressed breastmilk into foremilk and hindmilk. This evaluation will help us determine which fractionating method is more practical for mothers to use when separating expressed breastmilk. 1) Please assess the 3 fractionating methods on each of the following criteria - time to complete, convenience to you and clear & understandable directions (consider time for set-up, collection, separation and labeling). Using the scales below, assess each method by picking a number from the scale and recording it in the box provided for each method. a) How convenient did you find each method? C O N V E N I E N C E T O Y O U Visual Method 1 2 3 4 Time Method Volume Method inconvenient somewhat neither somewhat very inconvenient conv/inconv convenient convenient b) How clear and understandable did you find the directions for each method? C L E A R AND U N D E R S T A N D A B L E DIRECTIONS 1 2 3 4 5 Visual Method Time Method Volume Method unclear somewhat neither clear somewhat very unclear nor unclear clear clear c) How long (time) did each method take to complete compared to your usual expression time? Visual Method Time Method Volume Method TIME T O C O M P L E T E 2 3 significantly considerably moderately slightly no more time more time more time more time diff in time 150 Appendix 4 2) Please rank the 3 fractionating methods for overall preference. Use the ranking system provided and place the appropriate ranking number beside the method. Visual Method Time Method Volume Method Ranking system: 1 = most preferred method 2 = second preferred method 3 = least preferred method 3) Would you do any or all of the following methods for an extended period of time ( approximately 2 weeks) if required? Check the box that applies for each method. VISUAL METHOD TIME METHOD VOLUME METHOD Yes Yes Yes No No No Don't know Don't know Don't know 4) Do you feel there may be some instances when one method may be better suited than another method for a mother to use? Yes —> please describe: No Don't know 5) METHOD COMMENTS Visual Method lease provide any comments or suggestions on the fractionating methods Time Method Volume Method General Comments 151 SOCIO-DEMOGRAPHIC INFORMATION Appendix 4 The following section provides general information about the participants in the study. It is useful for researchers to know the general characteristics of the study group. The information will be kept confidential. Completion of this section is voluntary. If you do not wish to answer any of the questions, please feel free to leave the question blank. 1) Age: 2) Country of Birth: 3) How many years have you lived in Canada? 4) What city do you reside in? 5) What is your marital status? single (never married) married / common-law separated (but still married) divorced widowed 6) What is the highest level of education that you have completed? less than high school undergraduate degree high school graduate degree community college/technical school other: (specify: ) 7) What is your usual occupation? 8) What is your gross family income per year? less than $20,000 $20,000 - 50,000 > $50, 000 9) How many people are dependent on this income? 10) What is your ethnic background? Chinese Latin American European South Asian Middle East First Nations East Indian African Other:(specify ) 11a) How many children live in your household? b) What are the ages of the children? 152 Appendix 4 THANK-YOU FOR PARTICIPATING IN THIS STUDY YOUR TIME AND CONTRIBUTIONS ARE GREATLY APPRECIATED PLEASE RETURN THIS BOOKLET TO V IKK I LALARI WHEN COMPLETED ( A F I L E F O L D E R I S L O C A T E D A T T H E F R O N T D E S K I N S C N FOR D R O P O F F OR A S K Y O U R N U R S E T O C O N T A C T V I K K I T O P I C K - U P T H E B O O K L E T ) 153 Appendix 5 APPENDIX 5 Phase II Nutrition Study Booklet 154 Appendix 5 INSTRUCTIONS FOR SEPARATING, COLLECTING AND LABELING EXPRESSED BREASTMILK 1) Follow the SCN guidelines for set-up, handling and storage of expressed milk. 2) FRACTIONATING INSTRUCTIONS: follow the instructions provided for the method on how to separate milk into foremilk and hindmilk. 3) Label the containers as FOREMILK and HINDMILK. If more than one container is needed to collect Foremilk or Hindmilk, indicate the number of containers pumped in order of expression (i.e. Hindmilk #1/2, Hindmilk #2/2). If double pumping, you will have foremilk and hindmilk containers from each breast. 4) Using a syringe, take 1-2 mL sample of milk (from each container) of foremilk and hindmilk expressed during the 24 hr study period. These samples will be analyzed for fat content. (Note: if the sample is not taken immediately after pumping, shake the container gently to mix the fat before drawing milk up into the syringe) 5) Label the samples (syringes) as FOREMILK and HINDMILK with the labels provided. Fill in the following information: (Note: color coding system: blue = foremilk, yellow = hindmilk) DATE: TIME: NAME: R / L # / F O R E / H I N D MTHD: \ f f Circle right or left breast 6) Store samples in the refrigerator on the unit for the nutrition researcher to collect. 7) If pumping at home, refrigerate samples until you bring them to the hospital. Keep the samples cold for transporting by using a small "cooler chest" or thermal lunch bag with an icepack inside. 8) Once foremilk and hindmilk samples have been collected for the study, you can combine the milk into one container for storage and/or feeding. List the # Gradufeeds pumped <— (i.e. #1/2, #2/2) List the Method: Time method 156 BASELINE INFORMATION Appendix 5 A 24 hour milk expression diary is needed to collect baseline information on your usual pumping routine and volume of milk expressed. A sample of expressed milk will be collected during a 24 hour baseline period to analyze the fat content of your milk. * Use the table provided to record the baseline information. * Follow the " Instructions for Collecting, Storing and Labeling Expressed Breastmilk" on page 2. MILK EXPRESSION DIARY (Baseline) Date: TIME of Day DURATION of pumping session (minutes) TYPE OF PUMP (E=electric/ H =hand) METHOD of pumping (S= single D= double) SIDE of Breast (right=R, left=L, both=B) LOCATION of pumping session (home, hospital, other) VOLUME of milk expressed (per container per breast) mL Right 1 Left TOTAL VOLUME of milk expressed (mL) xample 8:00 am 20 mins electric double both hospital 30 mL 60 mL 90 mL 157 : Appendix 5 Please answer the following questions regarding breastmilk expression practices and possible factors that affect breastmilk supply. 1) List any factors that you feel may affect your milk expression routine. Eg. A busy home schedule makes it difficult for me to express regularly. 2) List any factors that you feel may affect your milk volume supply and indicate whether these factors increase or decrease your milk volume supply. Eg. I think pumping every 4 hours increases my milk supply. 3) List any factors that you feel may affect your "letdown" reflex (increased milk flow). Eg. Seeing my baby starts my letdown immediately. 4) From the list below, indicate if you do anything special to help stimulate your milk production/secretions. (You may check more than one box.) massage breast touching, seeing, hearing baby relaxation techniques pumping regularly reposition pump on breast no special techniques used thinking of your baby/picture of baby other: 5a) Does your expressed milk volume change from day to day? yes no -> go to #6 don't know - » go to #6 a) If yes, what is your range of milk volume expressed daily? Eg. 300-400 mL /day mL/day 158 Appendix 5 "Letdown" is an increased flow of milk that may occur during or before a pumping session? The following questions (#6 a,b,c) are about the "letdown" reflex. 6) Do you experience "letdown"? yes no -> go to #7 don't know^ go to #7 If you experience "letdown", please answer the following: a) When do you experience "letdown"? (if you experience more than one "letdown" during a pumping session, indicate by checking more than one box) before session beginning session middle session end session other: b) How often do you experience "letdown" when pumping? always ( all the time) most of the time ( more than half the time) sometimes (approx. half the time) occasionally (less than half the time) other: c) The following is a list of some signs/signals that "letdown" has occurred. Please indicate by checking which box best describes how you tell when "letdown " has occurred, (you may check more than one box) tingling sensation warm feeling "gushing" feeling increased volume other: 7) What is your daily fluid intake? (1 cup=8 oz) cups per day 8) Are you presently taking any prescribed medications that increase your breastmilk supply? yes no - » go to #9 don't know —> go to #9 If yes, please list the name of the medication: 159 Appendix 5 9) Are you presently taking any other products that specifically increase your breastmilk supply? Eg. herbs or herbal products yes no —> go to #10 don't know —> go to #10 If yes, please list the name of the product: 10) Are you presently taking any other prescription medications (other than those listed above) or any non-prescription medications? yes no-> go to #11 don't know -H> go to #11 If yes, please list: The following questions are about drinking caffeinated beverages (i.e. coffee, tea, cola-drinks) and herbal beverages. If you do not drink these beverages go to question #13. 11) In the past 4 weeks, have you consumed any caffeinated beverages? yes no —> go to #12 don't know -> go to #12 If yes, please answer the following: a) What type(s) of beverage did you consume? (check all that apply) coffee tea cola-type drink other: b) What is the average number of cups (1 cup=8 oz) you drink per week? cups 12) In the last 4 weeks, have you consumed any herbal beverages (i.e. herbal teas) or specialty teas? yes no -> go to #13 don't know ^ go to #13 If yes, please answer the following: a) List the types of herbal beverage or specialty tea. b) What is the average number of cups (1 cup=8 oz) you drink per week? cups 160 Appendix 5 The following questions are about alcohol use. If you do not drink alcohol please go to question #14. The term "drink" refers to 12 oz beer, 5 oz wine or 1.5 oz spirit/hard liquor. 13) In the past 4 weeks, have you taken a drink of alcohol? yes no - » go to #14 don't know - » go to #14 If yes, please answer the following: a) What is the average number of drinks consumed? per month b) What type of alcoholic beverage? (check all that apply) beer/cider/cooler wine spirits/hard liquor other: The following questions are about cigarette smoking. If you do not smoke cigarettes, go to question #15. 14a) At the present time, do you smoke cigarettes? yes no b) What is the average number of cigarettes you smoke per week? Many things can contribute to feelings of stress, including having a baby in the hospital. 15) Are you presently experiencing any stress? yes -> mild stress no -» go to #16 don't know -> go to #16 moderate stress severe stress b) If you are experiencing stress, do you feel this stress is affecting your i) milk supply yes no don't know ii) "letdown" reflex yes no don't know iii) milk expression routine yes no don't know 16) Please provide any additional comments on your breastmilk expression practices or factors you feel affect your breastmilk supply? 161 TIME (Letdown) METHOD Appendix 5 A milk diary ( 3 x 24hr) and sample of milk are needed to collect information on how to separate your expressed breastmilk into 2 fractions: foremilk (milk expressed at the beginning of a feed) and hindmilk (milk expressed for the remainder of the feed) based on the TIME (Letdown) METHOD. A small sample of foremilk and hindmilk will be collected to analyze the fat content. The TIME (Letdown) METHOD involves determining when "letdown" (increased milk flow) has occurred, after you start pumping. You will continue to pump for 2-3 minutes after "letdown". The milk expressed in this first fraction will be called "foremilk". Milk expressed after this point will be called "hindmilk". Example: FOREMILK- HINDMILK 2-3 mins LETDOWN Start of Pumping Point of separation (change gradufeed) End of Pumping INSTRUCTIONS: 1) Follow "Instructions for Separating, Collecting and Labeling Expressed Breastmilk" on page 2. 2) Use the TIME (Letdown) METHOD to separate expressed milk into foremilk and hindmilk. The "point of separation" will be 2-3 minutes after "letdown" occurs. (Note: the exact time will be confirmed with the researcher) You need a watch or clock close by to note the time. © If you do not experience "letdown", the point of separation will be 2-3 minutes after the start of pumping. (Note: the exact time will be confirmed with the researcher) 3) Change the container (gradufeed) at the "point of separation". The first fraction of milk expressed will be called FOREMILK. The remaining milk expressed will be called HINDMILK. 4) Fill out the table on the following pages for the Time Method. When recording the volume of milk expressed at each session, list which breast is pumped (R=right; L=left) and list the amount of milk collected from each breast. If more than one container is used to collect milk from a breast, indicate the order of expression by using numbers. For example: R1 (right breast, 1st container) and R2 (right breast 2nd container) C O M M E N T S : 162 Appendix 5 TIME (Letdown) METHOD Date: _ _ _ _ _ _ _ _ _ _ Weekday Weekend day Expressed Breastmilk will be separated at: minutes (after letdown/after startpumping) TIME of Day DURATION of pumping session (minutes) TYPE of pump (hand or electric) METHOD of pumping (single/ double) SIDE of breast (right=R, left=L, both=B) LOCATION of pumping session (home, hospital, other) LENGTH OF TIME to "Letdown" (minutes) VOLUME of milk expressed (list for each container) (R= right breast; L= left breast) FORE 1 HIND Example 8:00 am 20 mins Electric Double Both Hospital 2 mins R=20 mL L=30 mL R1=60mL R2=10mL L1=50 mL 163 Appendix 5 TIME (Letdown) METHOD Date: Weekday Weekend day Expressed Breastmilk will be separated at: minutes (after letdown/after startpumping) TIME of Day DURATION of pumping session (minutes) TYPE of pump (hand or electric) METHOD of pumping (single/ double) SIDE of breast (right=R, left=L, both=B) LOCATION of pumping session (home, hospital, other) LENGTH OF TIME to "Letdown" (minutes) VOLUME of milk expressed (list for each container) (R= right breast; L= left breast) FORE 1 HIND Example 8:00 am 20 mins Electric Double Both Hospital 2 mins R=20 mL L=30 mL R1=60 mL R2=10mL L1=50 mL 164 Appendix 5 TIME (Letdown) METHOD Date: Weekday Weekend day Expressed Breastmilk will be separated at: minutes (after letdown/after startpumping) TIME of Day DURATION of pumping session (minutes) TYPE of pump (hand or electric) METHOD of pumping (single/ double) SIDE of breast (right=R, left=L, both=B) LOCATION of pumping session (home, hospital, other) LENGTH OF TIME to "Letdown" (minutes) VOLUME of milk expressed (list for each container) (R= right breast; L= left breast) FORE 1 HIND Example 8:00 am 20 mins Electric Double Both Hospital 2 mins R=20 mL L=30 mL R1=60 mL R2=10mL L1=50 mL 165 Appendix 5 COMMENTS/NOTES: 166 Appendix 5 EVALUATION OF THE FRACTIONATING METHOD Your feedback is important in the development of a fractionating protocol for our nursery. Please evaluate the TIME method you have used to separate your expressed breastmilk into foremilk and hindmilk. 1) Please assess the fractionating method on each of the following criteria - time to complete, convenience to you and clear & understandable directions (consider time for set-up, collection, separation and labeling). Using the scales below, assess the method by picking a number from the scale and recording it in the box provided for the method. a) How convenient did you find the method? CONVENIENCE TO YOU Time Method I I 1 2 3 4 5 inconvenient somewhat neither somewhat very inconvenient conv/inconv convenient convenient b) How clear and understandable did you find the directions for the method? CLEAR AND UNDERSTANDABLE DIRECTIONS ,1 2 3 4 5 Time Method unclear somewhat neither clear somewhat very unclear nor unclear clear clear c) How long (time) did the method take to complete compared to your usual expression time? TIME TO COMPLETE Time Method significantly considerably moderately more time more time more time slightly no more diff time time 2) Would you do the TIME method for an extended period of time (approximately 2 weeks) if required? Check the box that applies for the method. TIME METHOD Yes No Don't know 167 SOCIO-DEMOGRAPHIC INFORMATION Appendix 5 The following section provides general information about the participants in the study. It is useful for researchers to know the general characteristics of the study group. The information will be kept confidential. Completion of this section is voluntary. If you do not wish to answer any of the questions, please feel free to leave the question blank. 1) Age: 2) Country of Birth: 3) How many years have you lived in Canada? 4) What city do you reside in? 5) What is your marital status? single (never married) married / common-law separated (but still married) divorced widowed 6) What is the highest level of education that you have completed? less than high school undergraduate degree high school graduate degree community college/technical school other: (specify: ) 7) What is your usual occupation? 8) What is your gross family income per year? less than $20,000 $20,000 - 50,000 > $50, 000 9) How many people are dependent on this income? (including your baby in the SCN) 10) What is your ethnic background? Chinese Latin American European South Asian Middle East First Nations East Indian African Other:(specify 11a) How many children live in your household? (including your baby in the SCN) b) What are the ages of the children? 168 Append ix 5 THAhJK-YOU Wn PARTICIPATING lb) THIS STUVY. YOUR TIMEAhlV CONTRIBUTIONS ARE GREATLY APPRECIATED PLEASE RETURN THIS BOOKLET TO VIKKI LALARI WHEN COMPLETED (A FILE F O L D E R IS L O C A T E D A T T H E F R O N T DESK IN S C N FOR DROPOFF OR ASK Y O U R NURSE T O C O N T A C T VIKKI T O PICK-UP T H E BOOKLET) 169 Appendix 5 Determination of "time" to separate milk into foremilk and hindmilk 1) Pump Time < 10 minutes (2min) 11-19 minutes > 20 minutes (3 min) 2) Letdown Yes —> before pumping beginning pumping @ min after start pump No Don't Know 3 ) Speed Milk Flow Slow ( < 25% total volume pumped within the first 5 minutes) (3 min) Steady Fast ( > 50% total volume pumped within the first 5 minutes) (2 min) 4) Milk Volume Production for fractionating low (<125%) (2 min) Average volume pumped= moderate (125 - 200%) Baby's intake = _mL/day _ml_/day high (>200%) (3min) % baby's intake=. 5) Right vs Left Breast Differences Volume —> _ _ _ _ _ _ predominant breast Speed of Milk flow-> breast has faster milk flow PLAN: SEPARATION TIME minutes = minutes + minutes Separation Time from Time Letdown Prescribed time start of pumping occurs after start for separation of pumping 0 mins before pumping Start of Pumping 10 min 20 min 5.min 15 min Adjustments: ± 0.5 min to 1 min (consider: hindmilk volume< baby's intake; large milk volumes) 170 Appendix 6 APPENDIX 6 Phase III Nutrition Study Booklet 171 T l O O CD C CO rz CD GO o co c C L cp_ 5' CD co cn CD I c TD z r 0) =5 C L 5' C Q fD zs C L C/) o JD CQ CD CD X TD —t CD W W CD Q . 0 cn £11 GO CD o CO < r= Q. 5 CD CD - • <^ 5, 5 ° cn _-. a- § o o _ ZS CD s< O c —* C/) o o o ' • CL CD 3 o CQ — \ fl) •a z r o' c r fD o CQ o rz zs C L fl) Z5 C L * < O zs^  aT Z5 «—i-c/T o z r D) -r fi) o i—t-CD co" o' cn O O 3 p_ 0 I—I-CD r o cn T J L- —! "D O TD_ < C L CD o -• o 2. -I ° 2. CD g -Ef =5 CD O (0 =5 _ i ^< CD Ej Z s co q -c r _ . ;?r CD —1 CD cn cn O Zi fl) o I—t-o" CD cn CD ZS C L QT O I-H-O —* cn Ef 0) i— 3 cu 0) —h —h CD O ^< O CD CD C L fl) Q . l—h CD CD CD TT C L fl) # ro Q . 0) CD 0 0 7C 0 Z5 C L C L fl) ^< 0 0) C/) OL CD —+> o -o J J o 0 ro < O wee CL' O —\ wee ed da kd; zs' 3. Si) *< Ef ?r cn cn' 0 CD c r X zs o TD C L okl res 0 CO weekend ion weekend diary nclud da 0 zs' —K O *< iform CO zs o fD zs o ' o o zs O cn ZS 0 yo cut rz —* <• 0 cz C/) C L cz fD fD_ << CO TO „—„ rz C L 3 fD "O cn zs' i—»• CQ z r fD. no. fD —^  zs' 0 0 ZS fD o ZS . C L ZS < fD O -^ lum ow) 0 o zs Ef 3 0 «—h fD 0 c r X 0 TD cn CD W CO 0 C L W L Example: 8:00 am TIME of Day 20 mins DURATION of pumping session (minutes) Electric TYPE of pump H=hand E= electric Double METHOD of pumping S=single D=double Both SIDE of breast R=right L=left B=both =c o CA • _ __ LOCATION of pumping session (hospital, home, other) JD 2J II II co o> o o 3 3 r - r-Right breast VOLUME of milk expressed (mL) per container, per breast L= 45 mL Left breast VOLUME of milk expressed (mL) per container, per breast 135 mL TOTAL Volume of milk expressed (mL) CD Q . CO CD CD TT CD Q. Q. CO CD m x •o 30 m co O z g > JO a > < 911 Example: 8:00 am TIME of Day 20 mins DURATION of pumping session (minutes) Electric TYPE of pump H=hand E= electric Double METHOD of pumping S=single D=double Both SIDE of breast R=right L=left B=both Hospital LOCATION of pumping session (hospital, home, other) 3 J 3 J ro -i II II CO o> o o 3 3 r - | -Right breast VOLUME of milk expressed (mL) per container, per breast L= 45 mL Left breast VOLUME of milk expressed (mL) per container, per breast 135 mL TOTAL Volume of milk expressed (mL) CD CD TT Q . CO *< CD CD CD CL CL CO O 0) CD 7s m x T> U m to co O z g > 33 < > < ro 9ZL Example: 8:00 am TIME of Day 20 mins DURATION of pumping session (minutes) Electric TYPE of pump H=hand E= electric Double METHOD of pumping S=single D=double Both SIDE of breast R=right L=left B=both Hospital LOCATION of pumping session (hospital, home, other) a u ro II II CO O J o o 3 3 |- r~ Right breast VOLUME of milk expressed (mL) per container, per breast L= 45 mL Left breast VOLUME of milk expressed (mL) per container, per breast 135 mL TOTAL Volume of milk expressed (mL) 3 ro S" 03 0 CO Z J CD Q _ Z J ' cL o o Si c ^ co= CO S3 zr = 2 CD ^ x CO " O 0 CD ^ co Q. « . 03 3 > < C Q CO ^ . 01 o 7S CZ 0 -~" C T zr CD O 0) S 3 Z T = : o ^ i i , os •< O CD cz -• C T CD cr g-o CZ —T cr I 03 ^ ! 03^ ^ - < ^ CD ^ <" CD co CD 03 03 Q . CD CO o —\ C T 0 O i C Q O O CD a o Z!_ i—i-?r Z J o C Q O o cn 03 D o o cz o CD 03 O zr 03 Z J C Q CD 5' o o o 03 Z J Q . O o o Z J CO co' CD Z J o o cz 0 03 CO Q . O cz -g zi' C Q CO 0 CO CO o' Z J 03 CO N =T. 03 O =T. Z J O Z J C T 03 O cr Z J ^ o-O g c to 0 O cr 03 cr 0 =s K CO Si OJ 5' c0 § ro ~ cr 3 §" I J 2 CZ 0 CO O Z J T 3 0 CO Z3 O n "a T 3 =J" 0 5 - 5 -2. C Q C Q Q3_ -~ ro » 0 C Q © O C CD I" _ C Q 0 CO cz CO 0 Q . 0 03 Z J ' C Q cr 03 cr III ^ T l Z< S 3 Er 03 0 — O _ zr ~ 0 cr o 0 o 3 S o "_. I* Er o 03 03 => CD £ ° ° rr < 0 0 ) Z J CO •o 0 o 0 o zr ZJ JZt ' cz 0 CO 0 TD CO 03. 0 *< o cz o Q -cz o I—t-o' Z J CO" 0 o —, 0 5' Z J CO co 03 CD ^ 2 0 ? q ,3 CO CO _ ra ro co 0 II Is o 0 r 03 2* 0 o 3 3 &) CD 0 . o K 3 m rz C Q CO — 03 >r 03 •a 2 c o 3 3 3' r r C Q Si 0 S 0 0 c: i s --&>. 0 o 3 c 03 55 "< 3' O 3 03 CO 0 CO 03 —»l 0 o < o c l i 3 3? ^ § 3- ^ > to 1 03 Z J Q . Q . o' Si 0 zr 0 Ef 0 0 CO 0 oT o o —^  CO 5' o 0 03 CO 0 C L 0 o —\ 0 03 CO 0 >< o cz < o 3 0 CO c C Q L7 to ^ 03 C 1^ o —^  CO ^ Er 2" CO 03 2. I - I S- 0 ^ _ ro _ Qj 3 (o W 03 O -=* 3 D 0 W 3 ~ CD 0 ^ CQ 3 3 5" ? 1 CO CO •a 3 (A CA 6' 3 CD 611 3 - ^ CO 3 -03 0 ) < CD —\ 03 C Q CD cr CD 3 T T cn O O CO c 3 CD Q . " D 0 03 CO 0 03 13 CO CD 0 3 C Q 0 CO C Q O O o o I T o co 3 " 0 T 3 03 CO 0 0 3 T CO 03 < 0 o c rt- Q> 03 ™ 0 03 Q _ 3 ' T 3 0 Q . O 3_ T T 3 o C Q O O * CD CD ; i . 5 3 5 <o « j . 0 c - i CD 01 o' O 3 N CA (A &) s. CD C CL "* a> c o O 3-^ 2. o 3 CA c 3 T3 =*. O 3 "5 c & 3 O 3 &> fr o 3-O &> CA CD IQ O •Q C CD CA o' 3 3 CD CD I 3 3 c? 3 co g. CO C Q i f C Q = CO - K 0 0 3 0 CO C Q O 3 3 o 03 S 3 Q . 2. < 3 ' g_ C Q c 3 0 o Ex 0 o 0 3 " S 0 I I 3 O i f ! w to o co' 8 03 C zr: 3 52. CD P- 2. o c CO o 3 0 3 CO » C Q O CO 0 O C Q 3 3 Si 3 " t &> S T 3 CD ° 8-3 O CD S D - =>. CO o o o c -^  —5 0 Q . T l 0 03 CO 0 5" Q . o' 03 0 C T o 3 " 0 O 2" 3 ' C Q 3 -o' 3 " o -o x CT 0 co Q . 0 co o —\ cr 0 CO 3 -o o cr 0 03 o 3 0 CO 3 3 5 W 3 CO Q3 0 ~* >< 1 O J O I 8 | a 0 ) 5- — 0 X I 0 CO ° ~ 3 S * 3 03 3 " O 3 O 5 3. =5; 0 0 3 ^ =K 0 9J rt- 3 ~~' 3 " 03 0 Zi rt. ^ 0 J. 0 3 0 I o S o CD 3 Q _ O • < O c 0 x 0 o 0 Q . o :> 3 " 0 3 " O C 3 •a 3 ' C Q C T 0 CO 0 CO CO o' 3 C T 0 C Q 3 ' 3 3 ' C Q CO 0 CO CO a. Q . 0 co 0 CO CO o' 3 0 3 C L CO 0 CO CO o' 3 0 § ? 8 o--o o 0 >< 3 . O 0 C 0 X ^ ® 3 - CD 03 £ => 0 o a o S 3_ C L a 22 3 ' C Q T 3 c 3 " O 3 ' C Q CO 0 CO CO o' 3 5' Q . o' 21 0 C T ^< O 3 " 0 O 3 ' C Q 3 o 0 03 3 O 3 0 C T O X •o c 0 X T 3 0 O 0 0 C L o •a 0 03 CO 0 03 3 CO 3 " 0 3 C Q 0 CO J\3 D o •< o c 0 x " O 0 —^ 0 ' 3 O 0 0 s o 3 o C Q O O co C L o 7KT 3 O C Q o co CD 3 o 03 CA CD a o o 3-&> o o o c & 3" CQ O ~i CP CD 5-&> T3 C I 3" <Q (A CD CA CA >»», o 3 SI CD o I 3" IQ •Q C CD CA C* o 3 (A 0) cr 0) Cr O c 8 0 3 3-CD 0 8 1 . 3 C T _ 3 g 2 CD 3 - D PT O co £. cz C L " O CD _ £D ZJ •o 03 C L C L Q3_ O o 3 3 CD O Z! o c CD DJ CO CD X -o —1 CD CO CO o' z> 03 o o' CD CO 0) o o —* CO •< o cz CD CD CD o I—* o c CD X T 3 —\ CD CO CO o cz CD CD a. o CD 0 X co c T 3 _ 0 CO 0 CO 0 CO ZJ o ZJ o ZJ o Q_ o ZJ. T T ZJ o C L O zs. ZJ o C L O ZJ o *< o cz 05 0 0 X " D 0 0 ' ZJ g ZJ' CQ CO —\ 0 CO cn C L O •< O c —K 0 0 ZJ" co' CO —* 0 CO CO co' CU, 0 o 5' CQ o 0 CO CO 0 < 0 3 _ O Q . 0 3 5 CD CO CO ^ - 0 CO 0 CO CO ~ W 0 CO CO ZJ o CQ o o =* 0 3 C L o ZJ o CQ o o o c 3 i? > 3 CD 5" CA 2 3 o o 3 5. cr c c? 0) 3 CD 0 CO 0 ZJ 0 X T 3 0 0 ' ZJ g ZJ' CQ 0) ZJ -< CO —t 0 CO CO CB CD 3 io (A o 3 CA y> 5" o c S: 3 <Q 3 -fi) s. 3 <Q a> cr a> cr 3 -CD 3 " o •8 £ > a 0 03 < 0 -1 03 CQ 0 ZJ c 3 cr 0 g CQ' 0> —t 0 i=t 0 CO o c CO 3 o 7 T 0 T 3 0 —\ s 0 0 TT •o 0 T 3 —I 0 CO 0 ZJ 0 Q . o *< o c CO 3 o 0 g C Q ' 03 -1 0 0 CO •o 0 CO ZJ o o 3 " CQ •Q c CD (A a: o 3 CA a> a> cr o c o <S" a> 3 CA 3 o 5-3 ' <Q c §• 3 o o o CQ' !? JA (Q o o •o c CD (A a: o 3 -4. cr 0 0 -^  g C L 0 -^ o o o_ 0 ZJ 0 CO X J T 3 0 o o o cr 0 < 0 —* 03 CQ 0 •o zr 0 o 03 Q . Si o 03 1.81. O O m z H Z81 ->• CD oT 3 - I Q) CD 5 03 CQ CD CO =r CD CD =j. Q_ C —T 3 =" CO CD 3" o Q_ 3 O c Q_ O c —K cr 0) cr *< CD CO o CO O o > CO 0)' 3 m 0) CO > CO 0>' 3 Q . 0 m 0) co o 03 OJ o 3 CO m c o •a 0 0) 3 o 0 CO •a 0 o 0) >< o c —* 0 E? 3 o' 0> o ?c CQ —T o c 3 Q . •o o 03 3 *< "O 0 o 0 o> —\ 0 C L 0 T 3 0 3 C L 0 O 3 3 O o 3 0 •O 3 O 3 CQ >< o —^ 0> CT <^ 3" 0 CO o oo 3" 0) CU SU CQ o o CO CO — H 03 3 3 o o o c c CO c SU_ o o o c T 3 03. 5' 3 •o 0 _ CO 3" CO CO 3" 3" § O 3 -C Q ' CO O O O CD T 3 0 *< 0 03 0 CO CO E? 03 3 ro o o o o o o 3 << t s 3. CD 03 CQ Q . 0 K. ~ 0 CD Q . 0 CQ 3 . o' 0 » 0 CO o 3 -o o ro o o o o cn o o o o 03 C L c 03 0 Q. 0 CQ T 0 0 Ol o o o o 0 CO •a 0 o 0 3 3" 03 3" 0 CQ 3" 0 CO 0 < 0 0 Q . C o o' 3 03 i—H O 03 < 0 o o 3 •a 0 0 C L •o CO. 0 s~ 0 < 0 0 C L co 0 T 3 03 0 Q_ CT 3 03 —* 0 ' C L < O o 0 Q_ C L O 0 C L 0 o. o o 3 3 o 3 o c —r 3 03 8> oT —t — CO oT CO •o •< o. o < o c —\ 0 CO C L 0 5' •o co X o :> 03 3 •< 0 03 3" 03 < 0 •< o < 0 Q_ o 03 3 03 O-03 •o ro O o 3 rt. —t o ro 3" > CQ 0 C C L O 3 o CO co H Cf 0 0 - * o o CQ O - a CQ CO - I CD S 0) A ' o 03 3 CO 0 03 3 •< o 0 . Q c 0 CO 6' 3 CO " O 0 03 CO 0 0 0 3 ' => O P , 3 § 03 5; n ° 3 » n ® 0 03_ rt! "S o 3 03 o § s 0 3 3 03 =r. cr 03 o ~ c o 5 § 0 CD g. 5" 2: 3 " O 9.1. — CO rt w ^ 3" CO =S 5 s • •y co 0 § ET CO Q . I| = OT § co 3 — c ^ " 03 CO • >< 2. 0 eo 0 03 0 —1 CO 3 o 0 CQ 0 3 0 03_ O 3 -03 —* 03 o CO O O o I o m o Q > TJ I oo > o o CO 3" 0 TJ TJ TJ CD CD CD CO CO CO tD (0 CD =3 =3 =5 t—h i—h i—*-5T r ? CD , CO <D 0 0 Q . CO CO o' CO. > C Q CD CQ CO to rjJ O rjj D Co i—I-CD i t CD O C Q CO > CQ CD CO t—h 3." co C Q - ^ CO 3 CO to CD O O | CQ 5^  O 3 CO. o O CO ZJ c r CD CD CL cr CD n o CJ T l O 33 CD CO CD CO —I o zr CD O £Z CL O O rr CO < CD CD (5 CO. o' Z> Appendix 7 APPENDIX 7 Conversion of Fatty Acid Content of Human Milk to Trigylcerides and Total Fat Content 185 Appendix 7 APPENDIX 7 Conversion of the Fatty Acid Content of Human Milk to Triglyceride (TG) and Total Fat Content The lipid in human milk consists primarily of triglycerides (98-99%), phospholipids (0.8%) and cholesterol (0.5%) (Jensen 1999, Bitman 1983). Different methods of lipid analyses can be used to determine the fat content of human milk. In the present study three methods of lipid analyses were used to determine the fat content of maternal milk: the modified Folch method, the Creamatocrit method and the Gas-Liquid Chromatography method. The modified Folch method used solvents such as chloroform and methanol to extract all the lipids in milk, including phospholipids and quantified the fat gravimetrically (Lammi-Keefe 1984). The Creamatocrit method determined a creamatocrit percent value that was used to calculate the total fat content of milk from the regression equation determined by Lucas et al (1978). The Gas -Liquid Chromatography method determined the fatty acid content of human milk by the analysis of fatty acid methyl esters. The fatty acid content of milk was not equivalent to the total fat content of milk derived by lipid extraction because it did not account for the glycerol fragment of the TG (Jensen 1999). A conversion factor was required to convert the fatty acid content of milk to TG content. This was based on the assumption that TG accounted for 98-99% of the total fat content in the milk. Once the fatty acid content was converted to TG another conversion factor was applied to determine the estimated total fat content of the milk (total fat). The molecular weight of selected TG and their corresponding fatty acids were used to determine the TG conversion factor. The molecular weight of a TG consists of the sum of the molecular weight of the 3 fatty acids and the molecular weight of the gylcerol component, minus the molecular weight of 3 H 20 molecules. The following table provides the structure and molecular weight of 7 fatty acids which are present in quantities of >2% of the total fatty acid content of preterm milk (Bitman 1983). Table A1: Molecular Weight of Selected Fatty Acids1 and Gylcerol in Human Milk TG component Formula Molecular Weight2 Glycerol C 3 H 8 O 3 92 Laurie Acid (C12:0) C12H24O2 200 Myristic Acid (C14:0) C14H28O2 228 Palmitic Acid (C16:0) C16H32O2 256 Stearic Acid (C18:0) C18H36O2 284 Palmitoleic Acid (C16:1) C16H30O2 254 Oleic Acid (C18:1) C18H34O2 282 Linoleic Acid (C18:2) C18H32O2 280 Data from Bitman et al (1983) for selected fatty acids that are present in preterm milk in quantities of > 2% of the total fatty acid content. 2Molecular weight of fatty acid and glycerol based on the molecular wt of carbon (C =12), oxygen (0=16), & hydrogen ( H=1). 186 Appendix 7 A mean TG conversion factor was calculated from data reported by Winter et al. (1993) for the most common triglyceride found in human milk. The investigators found 22 triglycerides (with specific fatty acid profiles) that were present in quantities in milk > 1% of the total triglyercide content (Winter 1993). Using these fatty acid profiles the molecular weight of select fatty acid combinations and triglycerides were calculated (Table A2). The following equation was used to determine the conversion factor (Jensen 1999): TG Conversion Factor = mol wt of TG mol wt of fatty acids Table A2: Molecular Weight (MWT) and TG Conversion Factor for Select Triglycerides in Human Milk Fatty Acid % of Total TG1 MWT of MWT of TG Combination (mol/100molTG) Fatty acids2 TG3 Conversion Factor4 16:0+18:1+18:1 11.8 820 858 1.046 16:0+18:1+18:2 10.0 818 856 1.046 16:0+16:0+18:1 4.4 794 832 1.048 18:1+18:1+18:1 3.3 844 882 1.045 14:0+16:0+18:1 3.3 766 804 1.050 16:0+18:2+18:2 3.2 816 854 1.047 12:0+16:0+18:1 3.1 738 776 1.051 16:0+18:0+18:1 3.1 822 860 1.046 14:0+18:1+18:1 2.8 792 830 1.048 16:0+16:0+18:2 2.4 792 830 1.048 14:0+18:1+18:2 2.3 790 828 1.048 16:0+18:1t+18:1 2.1 820 858 1.046 18:1+18:1+18:1 2.1 846 884 1.045 18:1+18:2+18:2 1.9 842 880 1.045 16:0+16:1+18:1 1.9 792 830 1.048 16:0+18:0+18:2 1.7 820 858 1.046 14:0+16:0+18:2 1.5 764 802 1.050 12:0+18:1+18:1 1.4 764 802 1.050 12:0+16:0+18:2 1.3 736 774 1.052 12:0+18:1+18:2 1.2 762 800 1.050 12:0+14:0+18:1 1.1 710 748 1.054 18:0+18:1+18:2 1.0 846 884 1.045 Mean 1.048 ii 1 1 1 1 ; — ; i Adapted from Jensen (1999) and Winters et al (1993). Amounts of selected TG >1 % of total TG in human milk. 2 MWT of fatty acid = sum of MWT of fatty acid sn1 + fatty acid sn2+ fatty acid sn3 3 MWT of TG = (sum of MWT of fatty acid sn1+ fatty acid sn2+ fatty acid sn3) + (MWT of glycerol) - (MWT 3 x H20) 4 T G conversion factor = MWT of the TG + MWT of the 3 fatty acids in the corresponding TG A mean conversion factor of 1.05, which was similar to that suggested by Jensen (1999) was used to calculate the TG values in this study. Another conversion factor, 1.013, was applied to the TG value to obtain the total fat content of milk. This conversion factor assumes that 98-99% of the total fat in milk is in the form of triglycerides and the remaining fat or lipid (1.3%) is present mainly as phospholipids and cholesterol (Bitman 1983). [The total fatty acid content of milk (g/1 OOmL) can be converted to total fat by applying a conversion factor of 1.063.] 187 Appendix 8 APPENDIX 8 Fresh EBM Creamatocrit Measurements: Relations Between Milk Fat Content (g/1 OOmL) versus Creamatocrit Value (%) 188 Appendix 8 APPENDIX 8 Fresh EBM Creamatocrit Measurements: Relations Between Milk Fat Content (g/1 OOmL) versus Creamatocrit Value (%) « — ~ T • 1 • ' I • 1 B 1 • 1 • 1 • 1 1 1 • I V 1 • 1 h 0 1 2 3 4 5 6 7 8 9 10 11 12 Creamatocrit (%) Value Figure A1 Relations Between Fat Content Values as Determined by the Modified Folch Method and the Creamatocrit Method1'2 1Results are show for expressed breast milk samples collected from 12 mothers in Phase I and Phase II, n=68. The creamatocrit value was determined for fresh milk samples by the Creamatocrit Method.The milk fat content (g/1 OOmL) was determined for frozen milk samples by the modified Folch Method. 2Solid lines represent the regression line (middle line) and 95% confidence interval (outer lines); r=0.928, P<0.001; regression equation, Y'=0.567X-0.159. A regression equation was developed to predict the estimated total fat content of milk from the creamatocrit value. The creamatocrit value was determined for fresh milk samples by the Creamatocrit Method. The milk fat content (g/1 OOmL) was determined for frozen milk samples by the modified Folch Method. The regression equation for fresh milk was Y' = 0.567X - 0.159, (R2=0.860, P<0.001, n = 68). 189 Appendix 9 APPENDIX 9 Phase I and Phase II: Combined Data for Milk Fat Content and Volume of Expressed Milk 190 Appendix 9 Appendix 9 Phase I and II: Combined Data for Milk Fat Content and Volume of Expressed Preterm Milk The Fat Content of Preterm Milk Data from Phase I and Phase II were combined for the baseline period to compare the results of total fat content of milk using three different methods of lipid analysis: the modified Folch method, the fatty acid method (conversion to total fat equivalents), and the Creamatocrit method (using the Lucas equation). Twenty-seven mothers of premature infants successfully completed the baseline period in the study, 19 mothers from Phase I and 8 mothers from Phase II. The quartile range for the fat content of preterm milk collected in Phase I and II was 3.26 g fat/1 OOmL for the 25 t h %ile, 3.83 g fat/1 OOmL for the 50 t h %ile, and 4.53 g fat/1 OOmL for the 75 t h %ile. Table A3 Determination of the Total Fat Content of Preterm Milk1 Mean ± SD Range Gravimetric Method, modified Folch method 3.86 ± 0.86 a 2.33-5.40 (g fat /100mL) TG Content 2 ' 3 3 .9710.89 2.21-5.71 (g fat /1 OOmL) Total Fat Content 2 , 4 4.02 ± 0.90 2.23-5.88 (g fat/1 OOmL) Creamtocrit Value 7.19 ± 1 33 4.33-9.17 (%) Creamatocrit Method, Lucas Equation 5 4.52 ± 0.92 a 2.56-5.88 (g/1 OOmL) Results are from the baseline period for Phase I and II, n=27. 2 Based on determination of total fatty acids in milk; 3.78 ± 0.84 g total fatty acid/1 OOmL (range 2.10-5.44). 3 T G (triglyceride) content of milk was determined by converting total fatty acids to TG, by applying the conversion factor 1.05. 4 Total fat content of milk was determined by converting total fatty acids to total fat by applying the conversion factor 1.063. 5 The regression equation developed by Lucas et al. (1978) for estimating the total fat content of milk: fat (g/L)= (creamatocrit value - 0.59 )/0.146. Significantly different using paired t-test P<0.001 A.9.2 Volume of Expressed Preterm Milk The mean volume of milk expressed by mothers during the 24-hour baseline period in Phase I and II (combined data) was 794 ±318 mL/day (mean ± SD), ranging from 293 -1435mL/day (n=28). 191 Appendix 10 APPENDIX 10 Medium Chain Fatty Acid Content of Preterm Milk 192 APPENDIX 10 Medium Chain Fatty Acid Content of Preterm Milk Appendix 10 Medium chain fatty acids, MCFA, are saturated fatty acids with chain lengths of 6-12 carbons (Bach 1982). As a result of their lower chain lengths and smaller molecular size, M C F A are more soluble in water than longer chain length fatty acids (Jensen 1992, Bach 1982, LePage 1984, Christee 1982). This property may lead to potential losses of MCFA in the aqueous phase of the modified Folch procedure (solvent extraction lipid analysis method) and thereby may cause an underestimation the total fat content of the milk, when determined by this method. To examine this possibility, the baseline milk samples from Phase I and II (n=27) were analyzed for total fatty acid and M C F A content. The mean medium chain fatty acid, MCFA, content of preterm milk from mothers (n=27) in Phase I and II during the baseline period was 0.26 ± 0.07 g/1 OOmL, range 0.17-0.43 g/1 OOmL. The mean percentage of M C F A relative to total fatty acid content of preterm milk was 7.3 ± 2.4% (range 4.6-16.6%). These values are comparable to values of M C F A (C8:0 - C12:0) content of preterm milk reported by other investigators. Ehrenkranz et al (1984) reported the mean M C F A content of preterm milk (n=14, day 28 postpartum) represented 4.92% of the total fatty acid content of milk. Lepage et al. (1984) reported the mean M C F A content of preterm milk was 10.8% of the total fatty acid content (n=7). Bitman et al. (1983) reported the mean M C F A content of preterm milk (day 42 postpartum) represented 7.82% of the total fatty acid content for premature infants at 31-36 weeks gestation (n=28) and 8.84% for premature infants at 26-31 weeks gestation (n=18). Table A4 Total Fatty Acid Content and MCFA content of Preterm Milk1 Mean ± SD Range Total Fatty Acid (g/1 OOmL) 3.78 ±0.84 2.1-5.44 M C F A 2 (g/1 OOmL) 0.26 ±0.07 0.17-0.43 % MCFA relative to total fatty acid content 7 27 ±2 4 4.5-16.6 (%) 1 Fatty acid content was determined by direct methylation (Lepage and Roy 1988) and gas chromatography; samples from baseline period of Phase I and II (n=27). 2The content of MCFA (medium chain fatty acids) are calculated as the sum of C8:0 + C10:0 + C12:0 fatty acids. 193 Appendix 11 APPENDIX 11 Phase I - Creamatocrit Measurements for Foremilk and Hindmilk 194 Appendix 11 Appendix 11 Phase I - Creamatocrit Measurements for Foremilk and Hindmilk Table A5 Creamatocrit Measurements for Foremilk and Hindmilk in Phase I (Completed Group)1 Foremilk Hindmilk Composite2'3 (%) (%) (%) Visual Method 4.85 ± 0.94 8.47 ± 0.83 6.60 ± 0 . 7 1 T ime Method 5 . 0 8 ± 1.21 8.80 ± 1.26 7.11 ± 0 . 9 2 Vo lume Method 5.50 ± 1.27 9.27 ± 1.89 7.09 ± 1.18 Values are given as mean ± SD creamatocrit % values for n=10 subjects who successfully completed Phase I. 2 Composi te milk value is calculated from the respective volumes and creamatocrit values of foremilk and hindmilk. 3 The baseline mean creamatocrit value for the completed group was 6.96 ± 1.78 %, for the completed group (n=10). 195 Appendix 12 APPENDIX 12 Expressed Breast Milk Practices Survey Results 196 Appendix 12 Appendix 12 EXPRESSED BREAST MILK PRACTICES SURVEY RESULTS Resul ts of the survey of expressed breast milk pract ices were combined for the three study phases . The survey explored potential factors that may affect the mothers' milk express ion routines and milk production. A total of 44 mothers completed the survey, 21 mothers in P h a s e I, 8 mothers in P h a s e II and 15 mothers in P h a s e III. Techniques Used to Stimulate Milk Production Forty-four mothers provided information on techniques they used to stimulate their milk production (Table A6). A variety of techniques were used by mothers to stimulate milk production, the most common techniques were: regular pumping, massag ing the breast, reposit ioning the pump on the breast and thinking of their baby. Most mothers (n=34/44, 77%) reported using more than one technique for stimulating milk production. Table A6 Techniques Mothers Used to Stimulate Milk Production1'2 n % Pumping regularly 34 77 Massage Breast 25 57 Reposition Pump 20 45 Thinking of baby, picture of baby 17 39 Touching, seeing, hearing baby 11 25 Relaxation Techniques 9 20 No special techniques 7 16 Other 3 10 23 Results are shown for mothers from all 3 study phases who completed the expressed milk practices survey, n=44; 21 mothers from Phase I, 8 mothers from Phase II, and 15 mothers from Phase III. 2Many mothers reported using more than one technique, thus the percentages do not totaHOO. 3Other techniques reported by mothers included: heat pack, hot shower, switch one breast to the other breast, take deep breath, different electric pump machines, eat pig hoof with peanut sauce, & warm cloth. The Letdown Reflex Table A7 descr ibes the mothers' exper iences with the letdown reflex. Forty of the 44 mothers (91%) who completed the survey reported exper iencing the letdown reflex. Of the 44 mothers who completed the survey, 2 mothers did not provide information on the timing of letdown, and 4 mothers did not provide information on how often they exper ienced letdown or the s igns and signals of letdown. Most of the mothers exper ienced letdown before a pumping sess ion (n=22/42) or at the beginning of a 197 Appendix 12 pumping sess ion (n=29/42). S o m e mothers (n=21/42) reported exper iencing letdown at different t imes during lactation, for example, at the beginning of a pumping sess ion and during the middle of a pumping sess ion . Most of the mothers (n=18/42) reported they "a lways" exper ienced letdown, while 11 of the 42 mothers reported letdown occurred "most of the time" (more than half of the time). Mothers had a variety of s igns and signals that indicated letdown was occurr ing. The most common signs or signals of the letdown reflex reported by the mothers were a tingling sensat ion (n=25/40) and an increased flow of milk (n=22/40). Table A7 Mothers' Experiences with the Letdown Reflex1,2 n % Experienced Letdown Yes 40 91 n=44 No 3 7 Don't know 1 2 Timing of Letdown 3 Before Pumping Session 22 52 n=42 Beginning Pumping Session 29 69 Middle Pumping Session 9 21 End Pumping Session 3 7 Other 4 2 5 How often experienced Always (all the time) 18 43 Letdown 5 Most time (more than half the time) 11 28 n=40 Sometime (approx. half the time) 7 18 Occasionally (less than half the time) 4 10 Other 6 2 5 Letdown signs and Tingling feeling 25 63 signals 7 "gushing" feeling 12 30 n=40 Warm feeling 10 25 Increased flow 22 55 Other 8 11 28 Results are shown for 44 mothers from who completed the milk expression survey from the three study phases: 21 mothers from Phase I, 8 mothers from Phase II, and 15 mothers from Phase III. Two mothers did not provide information on the timing of letdown and 4 mothers did not provide information on how often letdown was experienced or the signs & signals of letdown. 3Some mothers reported experiencing letdown at different times during lactation, thus the percentages shown do not total 100. 4Other: "after session; skin-to-skin cuddling; random pain or squeezing feeling in beasts usually accompanied by leaking". 5Some mothers reported experiencing letdown from more than one category, thus the percentages shown do not total 100. 6 Other: "never; especially when really full". 7Some mothers reported experiencing more than one type of letdown sign/signal, thus percentages shown do not total 100. 8Other:"pulling feeling in nipple; aching; breast full and hard; hurt feeling/tight feeling/pushing out; pressure; tensing feeling; cross between a tingling sensation and cramping; pain in breast; visually wet spots; no feeling felt; pain(squeezing or tightening)". 198 Appendix 12 Galactagogues and Milk Production A smal l number of mothers in the study were taking ga lactagogues, products to increase their expressed milk production. S ix of the 44 mothers (13%) who completed the expressed breast milk pract ices survey were taking Motil l ium ®, a prescr ibed medicat ion used to increase milk supply. Three of the 6 mothers taking Moti l l ium® were from P h a s e I, and 3 mothers were from P h a s e III. The volume of milk expressed by mothers taking Moti l l ium® ranged from 231-942 mL/day. The percentage of milk expressed by mothers relative to their own infant's intake ranged from 72-327%. Five of the 6 mothers were able to produce enough milk to meet their own infant's enteral intake. O n e mother was unable to produce enough milk to meet her infant's enteral intake. A few mothers reported taking specif ic products to increase their milk production. Two mothers reported taking the herb fenugreek to increase their milk supply. O n e mother reported eating a traditional East Indian food cal led pinne, (made of a lmonds, raisins, brown sugar, & flour) to enhance her milk production. Three mothers reported drinking Guinness (beer) to help increase their milk supply. Cigarette Smoking and Milk Production Forty-three mothers provided information on cigarette smoking. S e v e n of the 43 mothers (16%), 6 mothers from P h a s e I and 1 mother from P h a s e II, reported smoking cigarettes. The vo lume of milk expressed by mothers who smoked cigarettes ranged from 293-1435 mL/day. S ix of the 7 mothers produced over 500 mL of milk per day and 2 of the 7 mothers produced over 1000 mL of milk per day. Al l the mothers produced enough milk to meet their own infant's enteral intake. The percentage of milk vo lume expressed by mothers relative to their infant's intake ranged from 125-664% (n=7). Alcohol Consumption and Milk Production Forty-three of the 44 mothers who completed the expressed milk pract ices survey provided information on their a lcohol consumpt ion. Eighteen of the 43 (42%) mothers reported consuming an alcohol beverage within the last 4 weeks of participating in the study. Three mothers reported drinking Guinness (beer) to help increase their milk supply. The mean consumpt ion of alcohol ic beverages was 2.3 199 Appendix 12 drinks per month (0.6 drinks per week), where one drink was equivalent to 12 oz beer/cider, 5 oz wine, or 1.5 oz spirit/hard liquor. The range of alcohol ic beverage intake was 0.5-8 drinks per month or 0.13 to 2 drinks per week. The volume of milk expressed by mothers who consumed alcohol ranged from 231-1355 mL/day. The percentage of milk vo lume expressed by mothers relative to their infant's intake ranged from 72-1255% (n=18). Two mothers were unable to produce enough milk to meet their infant' intake. Stress and Milk Production All 44 mothers who completed the expressed milk pract ices survey reported they exper ienced some degree of stress (Table A8). Eighteen of the 44 mothers exper ienced a mild level of stress, 23 mothers reported exper iencing a moderate level of stress and 9 mothers reported exper iencing a severe level of stress. Severa l mothers (n=5/44) reported exper iencing more than one level of s t ress, for example , one mother reported exper iencing all levels of stress and commented that the degree of stress "depended on (her) baby's condition". Two mothers did not provide information on the affect of stress on their letdown reflex. O n e mother did not provide information on the affect of stress on her milk supply or her pumping routine. Seven teen of the 43 (40%) mothers felt stress affected their milk supply, 11 of the 42 mothers (28%) felt stress affected their letdown reflex and 22 of 43 mothers (50%) felt stress affected their pumping routine. The vo lume of milk expressed by mothers who reported exper iencing moderate or severe stress ranged from 231-1355 mL/day. The percentage of milk vo lume expressed by mothers relative to their infant's intake ranged from 72-1255%. One mother who reported exper iencing a moderate level of s t ress was unable to produce enough milk to meet her infant's intake. 200 Appendix 12 Table A8 Stress and Milk Production1'2 n % Experienced Stress 3 Yes, mild 18 41 n=44 Yes, moderate 23 52 Yes, severe 9 20 No 0 0 Stress affected milk supply Yes 17 40 n=43 No 15 35 Don't know 11 26 Stress affected letdown Yes 11 28 n=42 No 19 48 Don't know 12 25 Stress affected pumping routine Yes 22 50 n=43 No 16 36 Don't know 5 11 Results are shown for 44 mothers from who completed the milk expression survey from the three study phases: 21 mothers from Phase I, 8 mothers from Phase II, and 15 mothers from Phase III. 2One mother did not provide information on stress and milk supply or pumping routine and two mothers did not provide information on stress and letdown. 3Some mothers reported experiencing more than one level of stress, thus the percentages shown do not total 100. Visual Changes in Milk Observed by Mothers During Pumping Fifteen mothers who completed the survey in P h a s e III provided information on whether they observed a visual change in their milk during a pumping sess ion . S e v e n of 15 mothers (47%) noticed a change in the color and/or cons is tency of their milk during a pumping sess ion , while 8 mothers (53%) did not notice a change in the milk color and/or cons is tency of their milk. Se lec ted comments from mothers regarding their visual observat ions on the change in milk color and/or cons is tency are listed below. P 1 3 "The foremilk is a very thin and white and watery, then after letdown it is thicker, creamier and slightly yellowish color." P14 "Sometimes the milk gets creamier from the middle of a session to the end." P 1 6 "Thinner at beginning (watery) changes to a thicker consistency halfway through." 201 Appendix 12 Mother's Observations on Day-to-Day Variation of Milk Volume Forty-three of the 44 mothers who completed the expressed milk pract ices survey provided information on whether their expressed milk vo lume changed from day to day (daily variation). Twenty-eight mothers (65%) felt their milk vo lume changed from day to day, 11 mothers (26%) did not feel their milk vo lume changed from day to day and 2 mothers (5%) did not know if their milk vo lume changed from day to day. Factors Affecting Milk Expression Routine, Milk Supply and Letdown Reflex Forty-four mothers who completed the expressed breast milk pract ices survey provided addit ional information on factors they felt affected their milk express ion routine, their milk supply and their letdown reflex. T h e m e s were generated from the mothers' responses to the open-ended quest ions asked in the survey. The most prominent themes regarding factors affecting the mothers' milk express ion routine were: busy schedule (n=23), a c c e s s to pumps (n=10), traveling to hospital (n=8), pumping schedule (n=8), stress (n=7), baby's care/condit ion (n=6) and maternal fat igue/sleep (n=6). Se lec ted comments from mothers regarding factors they perceived as affecting their milk express ion routine are provided below. D11 "A busy schedule away from home makes it difficult for me to have access to a pump." P06 "Sometimes the pump room is being occupied by other SCN mothers, which more often than not puts me behind schedule." D09 "Stress is a major factor in being able to express breast milk." D14 "... planned pumping around traveling (have pump that has adapter to plug into car." D19 ".. .procedures for baby (be there first), pump secondary." V02 " If I get over tired... my milk doesn't come in as well." The most prominent themes generated regarding factors affecting the mothers' milk supply were: regular pumping (n=26), maternal fluid intake (n=20), pumping technique (n=15), maternal fat igue/sleep (n=10), maternal diet/eating (n=11) and stress (n=6). 202 Appendix 12 Se lec ted comments from mothers regarding factors they perceived as affecting their milk supply are provided below. V05 " Being regular on pumping will increase my milk supply." D22 " eating and drinking (if I eat less or drink less or miss a meal it affects supply)" D05 "... double pumping - increase milk supply; warm cloth/shower, increases." P11 "Pumping every 3 hours with a 6 hour sleep period increase my supply." D21 "Stress decreases milk supply." The most prominent themes generated regarding factors affecting the mothers' letdown reflex were: baby contact/thinking of baby (n=26), pumping technique/routine (n=16), maternal health/relaxation (n=7), and no factors (n=6). Se lec ted comments from mothers regarding factors they perceived as affecting their letdown reflex are provided below. D06 "Holding my baby causes the letdown reflex for me. Hearing or seeing helps as well." V04 " The sound of the breast pump can cause letdown..." P19 "massaging breasts... increase my letdown" P08 "Being overall comfortable and relaxed also helps" (with letdown) D13 "No letdown felt." 203 Appendix 13 APPENDIX 13 Comparison of Selected Regression Equations for Estimating the Total Fat Content of Human Milk (g/1 OOmL) for the Creamatocrit Value (%) 204 Appendix 13 Appendix 13 Table A9 Comparison of Selected Regression Equations for Estimating the Total Fat Content of Human Milk (g/ 100mL) from the Creamatocrit Value (%) Creamatocrit Value (%) Present Study Equation1 Lucas Equation2 Wang Equation3 Meier Equation4 1 0.39 0.28 0.93 0.43 1.5 0.68 0.62 1.20 0.68 2 0.96 0.97 1.47 0.93 2.5 1.25 1.31 1.74 1.18 3 1.54 1.65 2.01 1.43 3.5 1.82 1.99 2.28 1.68 4 2.11 2.34 2.55 1.93 4.5 2.39 2.68 2.82 2.19 5 2.68 3.02 3.09 2.44 5.5 2.97 3.36 3.36 2.69 6 3.25 3.71 3.63 2.94 6.5 3.54 4.05 3.90 3.19 7 3.82 4.39 4.17 3.44 7.5 4.11 4.73 4.44 3.69 8 4.40 5.08 4.71 3.94 8.5 4.68 5.41 4.98 4.19 9 4.97 5.76 5.25 4.44 9.5 5.25 6.10 5.52 4.70 10 5.54 6.45 5.79 4.95 10.5 5.82 6.79 6.06 5.20 11 6.11 7.13 6.33 5.45 11.5 6.40 7.47 6.60 5.70 12 6.68 7.82 6.87 5.95 12.5 6.97 8.16 7.14 6.20 13 7.26 8.50 7.41 6.45 13.5 7.54 8.84 7.68 6.70 14 7.83 9.18 7.95 6.95 14.5 8.11 9.53 8.22 7.21 15 8.40 9.87 8.49 7.46 15.5 8.69 10.21 8.76 7.71 16 8.97 10.55 9.03 7.96 16.5 9.26 10.90 9.30 8.21 17 9.54 11.24 9.57 8.46 17.5 9.83 11.58 9.84 8.71 18 10.11 11.92 10.11 8.96 1 1 1 Present Study equation (frozen samples): Fat (g/1 OOmL) = 0.572 (creamatocrit value) - 0.18 2 Lucas equation: Fat (g/L) = (creamatocrit value - 0.59) + 1.46 3 Wang equation (frozen samples): Fat (g/1 OOmL) = 0.54 (creamatocrit value) + 0.39 4 Meier equation: fat (g/1 OOmL) = 0.502 (creamatocrit value) - 0.074 205 

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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0090813/manifest

Comment

Related Items