ASSOCIATION BETWEEN ESSENTIAL FATTY ACIDS IN GROWTH AND COGNITIVE FUNCTION IN GHANAIAN CHILDREN By Mary Adjepong A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Human Nutrition - Doctor of Philosophy 2018 ABSTRACT ASSOCIATION BETWEEN ESSENTIAL FATTY ACIDS IN GROWTH AND COGNITIVE FUNCTION IN GHANAIAN CHILDREN By Mary Adjepong Background: There is a growing evidence from experimental and observational ]studies on the importance of essential fatty acid (n-3 and n-6 fatty acids) in growth and cognitive development of children. The dietary intake of fats in most developing countries is usually lower than adequate levels. In Ghanaian children below five years of age, 19% are stunted, 5% are wasted and 11% are underweight. Stunting levels have decreased in the overall population, however, the situation still persist in Northern Ghana where 33% of all under-fives are stunted. The reason for this disparity is unknown, however, interventions have focused on vitamins, minerals and proteins, but none has focused fatty acids (FAs). Recent studies have utilized fat based supplements, however, none of the studies have assessed the FA status in Ghanaian children neither have the relationship between FA status and cognition been measured. Therefore the objective of this research was to measure whole blood levels of FA in Ghanaian children and to establish the association of the FA levels with growth and cognitive abilities. Methods: The study was conducted in two populations in Ghana: Northern Ghana population - Savelugu-Nanton district (n=306) and Southern Ghana - Upper Manya Krobo district (n=209). A drop of whole blood was collected on an antioxidant treated card. The dried blood spot was analyzed for FA composition using gas chromatography. Weight and height were measured, age and gender information were taken. Weight-for-height, height-for-age, body mass index-for-age and weight-for-length z-scores were calculated with the World Health Organization Anthro software suite. Executive function was assessed with dimensional change card sort (DCCS) task. Seeds, nuts and oils were collected from locations in Northern and Southern Ghana. Fats were extracted from these foods by acidified methanol. FA composition of extracts was measured using the DSQII quadruple GC/MS. Minerals in seeds were quantified with ICP emission spectroscopy. Results: In the Northern Ghana population, 8.0% of children were essential FA deficient and 10.6% of the Southern Ghanaian children were essential FA deficient as defined by T/T ratio >0.02. In Northern Ghana: 29.7% of the children were stunted, n-6 FAs were inversely associated with stunting and n-3 FAs are positively associated with executive function. In Southern Ghana: 22.0% of children were stunted and no FAs was associated with stunting. When the FAs were compared for both regions, n-3 FAs levels were significantly higher in the Southern Ghana population and n-6 FAs level were significantly higher in the Northern Ghana population. Although these results are just associations, they indicate the possible role of n-6 FAs in growth and n-3 FAs in cognition. Melon seeds (agushie), soybeans, palm oil, fermented dawadawa, neri and cashew nuts were identified as good sources of EFA and minerals. Conclusion: This dissertation demonstrates a strong association between whole blood essential fatty acids and growth and cognition. It further confirms the essentiality of n- 6FAs in growth and gives insights on the relationship between n-3 FAs and cognition. Foods that contain EFAs in this population were also identified. These findings provide a basis of further research on how local foods can be used to eradicate malnutrition Copyright by MARY ADJEPONG 2018 This thesis is dedicated to my friend and partner the Holy Spirit. You brought me this far v ACKNOWLEDGEMENTS I will like to acknowledge God for the grace to accomplish this PhD program. Secondly, I will like to thank my family especially my brothers, Dennis, Anthony, Albert and Emmanuel for the emotional support and all the prayers. I will also like to thank my advisor and mentor Dr. Jenifer Fenton for her immense contribution in every step of the way: good times, tough times, all times, she was there! Dr. Suresh Babu, for being my mentor in the policy piece. You were a great help. Further, I will like to appreciate Dr. Matt Pontifex, Dr. Wei Li and Dr. Elizabeth Gardner, members of my committee as well as Dr. Reginald Adjetey Annan, my in-country coordinator, for their guidance and contribution. I appreciate the help of my collaborators, Dr. Grace Marquis of McGill university, Canada and Dr. Esi Colecraft of University of Ghana. To my mentors, Dr. Ibok Oduro and Dr. Karen Duca, I say a big thank you for your guidance. Also to Dr. Sarah Comstock, Dr. Theresia Jummane Jumbe and Dr. C. Austin Pickens for their help. To the undergraduates in my lab, Kelly Valentini, Jain Raghav and William Yakah, I say thank you. I would also like to thank my pastors, Bishop William Aggrey-Mensah, Lady Rev. Dr. Joy Bruce, Rev Raymond Nkrumah, Rev Gilbert Asamoah, and Ps Daniel Darku for all the support. I will also like to thank all my friends Mrs. Regina Adotey, Ms. Omolade Latona, Ms. Teopolina Suko Kanime, Mr. and Dr. Mrs. Vera Dzackah, Linda Stephens, Eunice France and Mr. Clement Kubuga, you guys are the best!! vi This material is based upon work supported by the United States Agency for International Development, as part of the Feed the Future initiative, under the CGIAR Fund, award number BFS-G-11-00002, and the predecessor fund the Food Security and Crisis Mitigation II grant, award number EEM-G-00-04-00013. Mary Adjepong is a fellow of the Norman E. Borlaug Leadership Enhancement in Agriculture Program funded by USAID. Support for this research was provided in part by the Borlaug Leadership Enhancement in Agriculture Program (Borlaug LEAP) (CGA#147309) through a grant to the University of California-Davis by the United States Agency for International Development The opinions expressed herein are those of the authors and do not necessarily reflect the views of the USAID. vii TABLE OF CONTENTS LIST OF TABLES ......................................................................................................... xii LIST OF FIGURES........................................................................................................ xv KEY TO SYMBOLS AND ABBREVIATIONS .............................................................. xvi CHAPTER 1: RATIONALE ............................................................................................. 1 Background ......................................................................................................... 1 Specific Aims ...................................................................................................... 2 Significance and innovation .............................................................................. 3 CHAPTER 2: LITERATURE REVIEW ............................................................................ 5 Introduction ......................................................................................................... 5 The Ghana story ................................................................................................. 7 Fatty acids ........................................................................................................... 8 EFA in Cognitive function ................................................................................ 10 EFA in Growth ................................................................................................... 11 Biomarkers of Essential fatty acid deficiency ................................................ 12 Research gap .................................................................................................... 14 CHAPTER 3: ASSOCIATION OF WHOLE BLOOD N-6 FATTY ACIDS WITH STUNTING IN 2-TO-6-YEAR-OLD NORTHERN GHANAIAN CHILDREN: A CROSS- SECTIONAL STUDY .................................................................................................... 16 Abstract ............................................................................................................. 16 Introduction ....................................................................................................... 17 Subjects and methods ..................................................................................... 19 Study setting ............................................................................................ 19 Ethical approval ....................................................................................... 20 Sample size and subjects ........................................................................ 21 Anthropometric measurements ................................................................ 21 Blood fatty acid assessment .................................................................... 21 Hemoglobin and malaria status ............................................................... 22 Data reduction and statistical analyses .................................................... 22 Results .............................................................................................................. 24 Subject characteristics ............................................................................. 24 Fatty acid levels in whole blood ............................................................... 26 Correlations between fatty acids and growth parameters ........................ 27 Relationships or associations between fatty acids and growth parameters ................................................................................................................. 29 Factor analysis ........................................................................................ 31 viii Discussion ........................................................................................................ 33 CHAPTER 4: WHOLE BLOOD N-3 FATTY ACIDS ARE ASSOCIATED WITH EXECUTIVE FUNCTION IN 2 TO 6-YEAR-OLD NORTHERN GHANAIAN CHILDREN. ................................................................................................................... 38 Abstract ............................................................................................................. 38 Introduction ....................................................................................................... 39 Methods ............................................................................................................. 41 Study site ................................................................................................. 41 Subjects and ethical approval .................................................................. 41 Anthropometric measurements ................................................................ 42 Blood fatty acid assessment .................................................................... 42 Hemoglobin and Malaria status ............................................................... 43 Cognitive assessment: Dimensional change card sort (DCCS) ............... 43 Data reduction and statistical analyses .................................................... 45 Results .............................................................................................................. 47 Subject characteristics ............................................................................. 47 FA levels in whole blood and regression analysis .................................... 49 Polytomous logistic regression ................................................................ 52 Factor analysis ......................................................................................... 53 Discussion ........................................................................................................ 55 CHAPTER 5: QUANTIFICATION OF FATTY ACID AND MINERAL LEVELS OF SELECTED SEEDS, NUTS AND OILS IN NORTHERN GHANA ................................ 60 Abstract ............................................................................................................. 60 Introduction ....................................................................................................... 61 Materials and Methods ..................................................................................... 63 Preparation of selected seeds, oils, and nuts in Ghana ........................... 63 Crude seed oil extraction ......................................................................... 64 Methylation of oils to FAMEs, neutralization, and FAME isolation ........... 65 FAME identification, analysis, and data Processing ................................ 65 Mineral analysis ....................................................................................... 66 Statistical Analyses .................................................................................. 66 Results .............................................................................................................. 67 Saturated fatty acids ................................................................................ 67 Monounsaturated fatty acids .................................................................... 69 Polyunsaturated fatty acids ...................................................................... 69 Minerals ................................................................................................... 72 Discussion ........................................................................................................ 74 Conclusions ...................................................................................................... 77 CHAPTER 6: ASSOCIATION OF WHOLE BLOOD FATTY ACIDS AND GROWTH IN GHANAIAN CHILDREN 2-6 YEARS OF AGE ............................................................. 78 Abstract ............................................................................................................. 78 ix Introduction ....................................................................................................... 78 Materials and Methods ..................................................................................... 81 Study setting ............................................................................................ 81 Sample size and subjects ........................................................................ 82 Ethical Standards Disclosure ................................................................... 82 Anthropometric measurements ................................................................ 83 Blood fatty acid assessment .................................................................... 83 Hemoglobin and malaria status ............................................................... 84 Dietary intake assessment ....................................................................... 84 Data reduction and statistical analyses .................................................... 85 Results .............................................................................................................. 86 Subject characteristics ............................................................................. 86 Fatty acid levels in whole blood ............................................................... 87 Relationships or associations between fatty acids and growth parameters ................................................................................................................. 90 Mean fatty acid levels for Northern and Southern Ghana, compared ...... 91 Dietary intake of proteins in Southern Ghana .......................................... 95 Discussion ........................................................................................................ 93 Conclusions ...................................................................................................... 97 CHAPTER 7: ASSOCIATION OF WHOLE BLOOD FATTY ACIDS WITH COGNITIVE FUNCTION IN 2 TO 6-YEAR-OLD SOUTHERN GHANAIAN CHILDREN. ................ 100 Abstract ........................................................................................................... 100 Introduction ....................................................................................................... 99 Methods ........................................................................................................... 102 Study setting .......................................................................................... 102 Sample size and subjects ...................................................................... 103 Ethical Standards Disclosure ................................................................. 103 Dietary intake assessment ..................................................................... 104 Anthropometric measurements .............................................................. 104 Blood fatty acid assessment .................................................................. 105 Hemoglobin and malaria status ............................................................. 105 Cognitive assessment: Dimensional change card sort (DCCS) ............. 106 Data reduction and statistical analyses .................................................. 107 Results ............................................................................................................ 109 Subject characteristics ........................................................................... 109 Regression between fatty acids and executive function measures ........ 113 Education and performance in the DCCS test ....................................... 118 Discussion ...................................................................................................... 119 CHAPTER 8: QUANTIFICATION OF FATTY ACID AND MINERAL LEVELS OF SELECTED SEEDS, NUTS, AND OILS IN GHANA................................................... 123 Abstract ........................................................................................................... 123 x Introduction ..................................................................................................... 124 Materials and Methods ................................................................................... 126 Preparation of selected seeds, oils, and nuts in Ghana. ........................ 126 Crude Seed Oil Extraction ..................................................................... 128 Methylation of Oils to FAMEs, Neutralization, and FAME Isolation ........ 128 FAME Identification, Analysis, and Data Processing ............................. 129 Mineral analysis ..................................................................................... 130 Results ............................................................................................................ 131 Saturated Fats ....................................................................................... 131 Monounsaturated Fatty Acids ................................................................ 131 n-3 Fatty Acids ....................................................................................... 134 n-6 Fatty Acids ....................................................................................... 134 Minerals ................................................................................................. 137 Discussion ...................................................................................................... 140 Conclusion ...................................................................................................... 144 CHAPTER 9: SUMMARY ........................................................................................... 145 Main findings .................................................................................................. 145 Innovation ....................................................................................................... 147 Limitations ...................................................................................................... 147 Future Directions ............................................................................................ 149 APPENDICES ............................................................................................................. 151 APPENDIX 1: Supplementary material ......................................................... 152 APPENDIX 2: Script to the consent form...................................................... 159 APPENDIX 3: Participant information collection sheet .............................. 166 APPENDIX 4: Policy component ................................................................... 168 APPENDIX 5: IRB approval letters ................................................................ 226 BIBLIOGRAPHY ......................................................................................................... 235 xi LIST OF TABLES Table 1: Sex differences are not associated with characteristics of participants ........... 25 Table 2: Nutrition and growth status of the children ...................................................... 25 Table 3: Median (Q1,Q3) of selected fatty acid proportions in whole blooda ................. 27 Table 4: Regression results between HAZ, WAZ and selected fatty acids .................... 30 Table 5: HAZ and WAZ regressed on calculated factors ............................................... 32 Table 6: Factor analysis of fatty acidsa .......................................................................... 32 Table 7:Characteristics of children who attempted the DCCS test ................................ 48 Table 8: Characteristics of children stratified by dimensional change card sort performance for the initial condition (Mean values and standard deviations; numbers and percentages) .......................................................................................................... 49 Table 9: 1Whole blood fatty acid proportions in Ghanaian children (Mean ± SE, n=307). ...................................................................................................................................... 51 Table 10: Regression results for performance on the dimensional change card sort test and selected fatty acids (FA). ........................................................................................ 52 Table 11: Associations of significant fatty acids, as tertiles, with dimensional change card sort test performance. ........................................................................................... 53 Table 12: Factor analysis of fatty acids.a ....................................................................... 54 Table 13: Regression† results for performance on the dimensional change card sort test and fatty acid (FA) factors ............................................................................................. 55 Table 14: Foods analyzed ............................................................................................. 67 Table 15: Saturated fatty acids expressed in mg FA/g crude oil or mg FA/g seed (Mean ± SD). ............................................................................................................................ 68 Table 16: Monounsaturated fatty acids expressed in mg FA/g crude oil, mg FA/g nut or mg FA/g seed (Mean ± SD) ........................................................................................... 70 Table 17: Polyunsaturated fatty acids expressed in mg FA/g crude oil or mg FA/g seed (Mean ± SD). ................................................................................................................. 71 xii Table 18: Mineral levels of seeds and nuts expressed as mg/kg .................................. 73 Table 19:Demographic characteristics of participants ................................................... 87 Table 20: Nutrition and growth status of children. ......................................................... 87 Table 21: Mean fatty acid levels for overall, stunted and non-stunted children in Southern Ghana (Values expressed as mean ± standard deviation) ............................ 88 Table 22: Regression results between HAZ, WAZ, and selected fatty acids ................. 89 Table 23: Regression results between WHZ, BAZ, and selected fatty acids (model: WHZ=fatty acid + hemoglobin + malaria; BAZ= fatty acid + hemoglobin + malaria) ..... 91 Table 24: Mean FA levels for Northern and Southern Ghana, expressed as mean ± standard deviation. ........................................................................................................ 94 Table 25: Characteristics of children who passed the initial instruction (n=24) ........... 111 Table 26: Characteristics of children stratified by dimensional change card sort performance for the initial condition (Mean values and standard deviations; numbers and percentages) ........................................................................................................ 112 Table 27: Regression results for performance on the dimensional change card sort test and selected fatty acids (FA). (Model: Total pass= Fatty acid of All Children + Age + BAZ + Hemoglobin+malaria) ....................................................................................... 114 Table 28: Regression results for performance on the dimensional change card sort test and selected fatty acids (FA). (Model: Total pass= Fatty acid of All Children + Age + BAZ + Hemoglobin+malaria) ....................................................................................... 116 Table 29: Regression results for performance on the dimensional change card sort test and selected fatty acids ............................................................................................... 118 Table 30: Educational component ............................................................................... 119 Table 31: List of seeds, nuts, and oils analyzed .......................................................... 127 Table 32: Saturated fatty acids (mg FA/g crude oil or mg FA/g seed)1 ........................ 132 Table 33:Total monounsaturated fatty acids (mg FA/g crude oil or mg FA/g seed)1 ... 133 Table 34: Total n-3 fatty acids (mg FA/g crude oil or mg FA/g seed)1 ......................... 135 Table 35:Total n-6 FA (mg/g crude oil of mg FA/g seed)1 ........................................... 136 xiii Table 36, S1: Median (Q1, Q3) of fatty acid proportions in whole blood ..................... 152 Table 37, S2: Regression results between HAZ, WAZ, and fatty acids ....................... 154 Table 38,S3: Spearman correlation (n=23) ................................................................. 155 Table 39, S4: Pearson correlation, n=23 ..................................................................... 156 Table 40, S5: Pearson correlation, n=38 ..................................................................... 157 Table 41, S6: Spearman correlation, n=38 .................................................................. 158 xiv LIST OF FIGURES Figure 1: Fatty acid biosynthesis in humans ................................................................. 10 Figure 2: Spearman correlations between participant characteristics, anthropometric measurements, and selected blood FA levels. .............................................................. 28 Figure 3: Illustration of the target cards and test cards used during the pre-switch and post switch .................................................................................................................... 45 Figure 4: Concentration of potassium (A), phosphorus (B) and magnesium (D) expressed as mg/kg .................................................................................................... 138 Figure 5: Concentration of iron (A), zinc (B), manganese (C) and copper (D) expressed as mg/kg. ..................................................................................................................... 139 Figure 6: Concentration of Calcium (A),Potassium (B), expressed as mg/g ............... 139 xv KEY TO SYMBOLS AND ABBREVIATIONS AA ALA BAZ Arachidonic acid Alpha linolenic acid BMI for age z-score DBS Dried Blood spot DCCS Dimensional change card sort DHA Docoxahexanoic acid EFA Essential fatty acids EFAD Essential fatty acid deficiency EPA Eicosapentanoic acid FA Fatty acids FADS Fatty acid Desaturase FAMES Fatty acid methyl esters GC-MS Gas Chromatography - Mass Spectroscopy HAZ LA Height for age z-scores Linoleic acid LCPUFA Long Chain Poly Unsaturated Fatty acid MUFA Monounsaturated fatty acid PUFA Poly Unsaturated fatty acid WHO World Health Organization xvi Background CHAPTER 1: RATIONALE During infancy and childhood, there is an increased need for nutrients in individuals because rapid growth and development occurs during this stage. This increased need causes infants and children to be susceptible to nutrient deficiencies. Nutrient deficiencies during this stage of life can have detrimental impacts among which is stunting. Stunting levels have declined in the past decades in Ghana but the situation still persist in Northern Ghana. Though the reasons for these regional differences in Ghana are not entirely clear, it is likely that location or local culture may govern limited access as well as cooking methods of the foods that could provide these dietary sources of fatty acids. The intake of fatty acids in most developing countries is low [1] and Ghana is not an exception and this dietary pattern could be leading to stunting and cognitive impairment. However, Ghana is home to seeds, nuts and oils that are high in fat but the EFA composition of some of these foods is unknown. This knowledge deficit may lead to under-utilization of such foods especially in child feeding. In an attempt to solve these problems in Ghana, a lot of interventions have been explored including the use of lipid based supplements [2], but the interventions do not assess the associations of various fatty acids in growth and cognitive function [3]. The supplements also utilizes only peanuts and soybean oil[2] but there are several seeds, nuts and oils in Ghana that could be explored and lead to diversification of the food supplementation. Apart from the supplements, the EFA rich foods identified can be included in child feeding. Considering the prevalence of stunting in Ghanaian children, the role of EFA in growth and cognitive function and the probable 1 under-utilization of EFA-rich seeds, nuts and oils, there is a need to know if there exist an association between EFA status and growth or cognitive function in children as well as the EFA composition of commonly consumed foods in Ghana. The long term goal of this study is to determine the physiological role of EFA in growth and cognitive function in Ghanaian children 2 to 6 years of age and also to identify some Ghanaian foods that are rich in EFA that can be used for probable interventions especially in child feeding. The short term goal is to assess EFA status in Ghanaian children and to know their association with growth and cognitive function as well as quantify the EFA composition of commonly consumed seeds, nuts and oils in Ghana. Our central hypothesis is that EFA levels is low in Ghanaian children between 2 to 6 years of age and EFA and will correlate with growth (as indicated by weight for age (underweight), weight for height (wasting), height for age (stunting) and BMI for age and cognitive abilities of these children. We will test the hypothesis using the following specific aims Specific Aims Specific Aim 1.Determine whole blood FA status in children 2-6 years of age from Northern and Southern Ghana and determine EFAD prevalence. Our central hypothesis is that there will be variations in EFA status in the different populations. Specific Aim 2a. Establish the associations between FA status and weight for height (wasting), weight for age (underweight), and height for age (stunting) and BMI for age z- scores in these children. Our central hypothesis is that whole blood EFA and z-scores will be positively correlated. It is further hypothesized that mead acid content and T/T ratio will be negatively correlated with the z-scores. 2 Specific Aim 2b. Assess the relationship between FA status and cognitive abilities of these children. Our central hypothesis was that EPA and DHA will correlate positively with cognitive function. We also hypothesized that Mead acid and T/T ratio will correlate negatively with cognitive function. Specific Aim 3. Identify oils, seeds and nuts used in this population and determine their FA and mineral content. Our central hypothesis is that some of the oil, seeds and nuts in Ghana have varying contents of EFA and minerals. Significance and innovation Though there have been studies in Ghana about EFA and child growth, these studies administered lipid based supplements without a detailed profiling of blood fatty acids [1], but this study will profile 24 fatty acids and know how the physiological changes of these fatty acids affect growth and cognitive function in children. The lipid based supplement adapted in a study in Ghana utilized only peanuts and soybean oil in its formulation [2], but there are seeds, nuts, and oils in Ghana whose EFA composition is unknown and the identification of such foods can lead to a diversification of the supplement formulation. Also the identification of these foods will be beneficial as these foods can be promoted to mothers so that they can be included in the foods for children. The EFA composition of the seeds, nuts and oils will be included in the West African food composition table. Moreover, the assessment of fatty acids have been ignored in developing countries due to technical (FA oxidation) and logistical (storage/shipment) problems. The collaboration with Dr. Harris, a leading expert in blood fatty acid analysis and the utilization of blood spot card pre-treated with an anti-oxidant cocktail (OxystopTM) for use in this 3 study is very innovative. This treatment assures sample integrity at room temperature for up to two weeks during storage and shipment. This ensures that samples arrive in the United States for a detailed fatty acid analysis which would be difficult to be conducted in Ghana. 4 Introduction CHAPTER 2: LITERATURE REVIEW Stunting and cognitive impairment is a global public health problem because it prevents people from realizing their mental and physical potentials [3]. Apart from the funds that are drained from various national resources to alleviate the problem, undernutrition in early life can lead to neurological damage that can lower the wages of individuals because they never finish their schooling. Stunting, which can be a result of poor fetal growth and sustained poor growth in early life, followed by rapid weight gain in later years, may predispose an individual to the development of chronic diseases such as diabetes and cardiovascular complications[4]. The causes of stunting and cognitive impairment include social, economic and cultural factors, maternal and child care, health and sanitation, household food security and nutrient intake. Among these causes, nutrient intake is the immediate cause because there are certain critical and sensitive stages during which adequate nutrition is essential [5]. This means that the diet of individuals should comprise all nutrients in their right proportions: carbohydrates, proteins, fats, vitamins, and minerals. The intake of fatty acids is important in growth and cognitive function in particular intakes of essential fatty acids (EFA) may prevent stunting and cognitive impairment. These EFAs include omega-3 fatty acids such as alpha linolenic acid and omega-6 fatty acids such as linoleic acid. Omega-3 fatty acids are found in foods such as flaxseed, soy, canola and walnut, while omega-6 fatty acid are found in foods such as sunflower and peanut oil. 5 These are referred as ‘essential’ because they cannot be synthesized in the body from metabolic precursors but need to be obtained from the diet [6]. EFAs (ALA and LA) and their derivatives (EPA, DHA and AA) have key roles in many developmental processes in humans. EFAs are responsible for myelination of brain neurons and the formation of synapses [7]. They are also known to have roles in stem cell proliferation and differentiation [8]. Supplementation of fish oils positively affects visual evoked potential [9], improves IQ maturation and visual acuity [10] in full term babies. EFA also improves verbal learning [11], short memory and linear growth in infants [12]. Progress in myelination and motor development [13] as well as faster processing of information [14] also occurs with intake of EFA in children. Improved problem solving [15] and better recognition memory [16] have also been documented in children who were supplemented with EFA. Additionally, maternal intake of EFAs is known to influence brain development of the fetus [17]. These are clear indications that EFAs can prevent stunting and cognitive impairment. In Ghana, intake of fatty acids in infants is below adequate levels. This may make children vulnerable to essential fatty acid deficiency (EFAD), leading to growth deficits and cognitive impairment. The main cause of this problem is poverty and food insecurity. Therefore establishing the relationship between EFA and growth and cognitive function is important. Assessing EFA content in seeds, nuts and oils will set the foundation for intervention using locally available foods to curb the situation. 6 The Ghana story Ghana is a small coastal country in West Africa with the majority of people residing the urban population. Like other developing countries, it has numerous natural resources but poverty still remains a problem especially in the north. Although agriculture in Ghana is a major contributor to the economy, it still remains at the traditional level where most farmers practice subsistence farming[18]. The Ghanaian diet is comprised of starchy roots (cassava, yam, and cocoyam), fruits (plantain, oranges) and cereals (maize, millet). Ghana is also the home of various seeds, nuts and oils that are rich in fatty acids. Among these are melon seeds, shea butter, tiger nuts, palm oil and coconut. The starches, cereals and fruits supply about 75% of dietary energy but proteins and lipids are below adequate levels. Due to urbanization, there has been a modification of the diet with an inclusion of wheat and rice but food insecurity remains a problem. [18] Breastfeeding is a common practice in Ghana however only about 50% of children below 6 months are exclusively breastfed [18]. After 6 months of exclusive breastfeeding, complementary foods are introduced to children which are sometimes low in nutrients. For example, it is reported that among children between 6 and 24 months, only 36.6% have some fats added to their complementary food [19]. These feeding practices, in addition to the composition of the Ghanaian diet, are among the causes of stunted growth in Ghanaian children. For children below age 5, the prevalence of stunting, wasting, and underweight is 19%, 5% and 11% respectively. Stunting has declined nationwide but the situation persists in the Northern Region with 33.1% of all children below age 5 in the region are stunted [20]. There is a need to know the cause of these regional variations. 7 The UNICEF Global Damage Assessment Report states that some children might not be realizing their physical, mental and intellectual capabilities due to nutrient deficiencies [3]. Apart from stunting, cognitive impairment may also occur due to nutrient deficiencies. Nutrients such as iron and iodine have been associated with cognitive impairment and stunting [21] but there is no published work linking EFAD to cognitive function or growth in Ghana despite the known crucial role of EFA in development. In an attempt to solve these problems, currently, many interventions have been explored including the use of lipid-based supplements [2], but the interventions do not assess the associations between various fatty acids and growth and cognitive function [1]. There are also several seeds, nuts and oils locally grown in Ghana that are nutritionally dense and provides essential nutrients but have never been considered as components of the child’s diet. These foods could be explored and included in child feeding as well as in food supplements. Considering the prevalence of stunting in Ghanaian children and its regional variation, there is a need to know the prevalence of EFA status of children in both regions, north and south, and how that relates to the prevalence of stunting. Also, given the role of EFA in growth and cognitive function it will be great to know the physiological role of individual fatty acids. Also, with the probable under-utilization of EFA-rich seeds, nuts and oils, EFA composition of commonly consumed foods in Ghana can be identified and assessed. Fatty acids Fatty acids play crucial roles in metabolism and these include their role as a major metabolic fuel (storage and transport of energy). They act as important components of 8 cell membranes in addition to their role in gene regulation. [22]. Beside these roles, some of them are closely associated with growth and cognitive development. Dietary fatty acids comprise of all lipids that are abundant in animal and plant tissues and are eaten as food. The commonest fats have been grouped based on their degree of saturation: saturated fats have no double bonds, monounsaturated fatty acids (MUFA) have a single double bond and polyunsaturated fatty acids (PUFAs) have two or more double bonds. Both saturated fats and unsaturated fats are further classified based on their chain length[6] Alpha linolenic acid (ALA) and linoleic acid (LA) are essential fatty acids, this means that, they cannot be synthesized in the body but must be included in the diet. EPA and DHA can be synthesized from ALA and AA can be synthesized from LA through desaturation, chain elongation and chain shortening (figure 1). The enzyme delta desaturase is an important enzyme in this. Both conversions from LA to AA and ALA to DHA compete for this enzyme and the affinity of the enzyme to substrate is in this order ALA>>LA>>OA. This has been used to develop biomarkers for EFAD[23]. Adequate intake of LA and ALA is based on the median intakes by different populations with diverse life stages and gender groups in the United States but there is insufficient data to set upper limits. Also, the term Adequate Intake (AI) is used because there is insufficient data to set RDAs. Consumption below the AI may lead to deficiencies in individuals and may have the capacity to cause chronic deficiencies[24]. 9 Figure 1: Fatty acid biosynthesis in humans EFA in Cognitive function It has been established that, EFAs play important roles in human development, both in fetal and neonatal development. Also the brain and retinal tissues have been shown to be highly dependent on EFAs, especially DHA for membrane fluidity and signal transduction. It has also been shown that, in children, EFAs contributes to cognitive development and may have a role in metabolic programming as well as in bone turnover and adipogenesis[25]. They also have roles in neurophysiology and disease prevention in adults [26] . These show that, throughout the life cycle, EFAs have crucial roles. 10 The period of growth and metabolic turnover in the entire human life cycle occurs during fetal developmental. It has been reported that, the average human brain exhibits a 60- fold increase in weight, from second trimester to two years; 20 to 1200g. Also, this is the period where neurons in the brain undergo rapid myelination and the synapses mature. It has been shown that ALA and LA levels decrease in the infant’s brain during the third trimester of pregnancy, but the levels of DHA and AA increases about 30-fold. An increase in concentration may be the body’s natural way of maintaining the right amounts for proper functioning. Among the functions of DHA and AA include the optimization of fluidity of cell membranes, modulation of neurotransmitter physiology including signal transduction by acetylcholine, dopamine and serotonin. They also have implications on cognition, vision and behavior. It has also been proven in randomized controlled trials in rats that, the groups fed with fish oils, soybean oil and safflower oil, had increased DHA and AA in the neuronal growth cone membrane. This suggests that, the maternal intake of dietary fatty acid may affect growth of fetal neurons and the biochemistry of neurotransmitters. Improved visual contrast discrimination, reversal task learning and psychomotor performance was enhanced in beagle puppies, which were fed with diets rich in DHA [22]. EFA in Growth Although there is little literature about EFA and children above 2 years, it is known that, an adequate amount and balanced supply of DHA and AA are required for growth and cognitive development. It has also been shown that the supplementation of the diet of children aged 7 to 9 year with bread spread rich in omega-3 showed an improvement in 11 verbal and learning ability of the children. In addition, supplementation of fish oils for children aged 3 to 13 years showed improvements in cognitive function [22]. EFAs may have a role in growth and scientifically based evidence indicates that there is an association with omega-3 PUFA intake and weight gain as well as height [27]. Though there are fewer studies to show that essential fatty acid deficiency (EFAD) influences protein energy malnutrition (PEM), there exists an association between EFA status and PEM. Though this claim has not been widely substantiated, both PEM and EFAD have similar symptoms, notably skin changes, impaired immunity, impaired and disturbed growth. It has been hypothesized that, a low fat intake can lead to impaired lipid digestion, absorption and transport and this can consequentially lead to desaturation and EFA beta oxidation and peroxidation. Prolonged EFAD may decrease lipid absorption hence worsens PEM by impairing nutrient absorption and dietary calorie utilization. Also dietary fats are known to provide energy for individuals, hence a diet that is deficient in fats can lead to PEM [28]. Biomarkers of Essential fatty acid deficiency EFAD is prevalent in several populations in the world and it is the result of the deficiency of alpha linolenic acid and linoleic acid; fatty acids that belong to the omega-3 and omega- 6 families. EFAD that is related to omega-3 deficiency is common and it can lead to impaired growth, skin lesions, infertility, kidney abnormalities, fatty liver disease, polydipsia and increased susceptibility to infections, as well as reduced learning and impaired vision. These symptoms may not be specific as they can be caused by other 12 clinical complications, hence an assessment of clinical symptoms as well as chemical assessment of imbalances between omega-3, omega-6, omega-7 and omega-9 fatty acids are used to diagnose EFAD [29]. Mead acid is regarded as the functional marker for EFAD. This is because, in the biochemical conversions of ALA to EPA and DHA, the enzyme responsible for the conversion faces competition from other substrates such as oleic acid and linoleic acid. The enzyme ∆- desaturase has an affinity to these substrates in this manner omega-3 >omega-6> oleic acid. Mead acid can only be formed from oleic acid when there is a deficiency of both omega-3 and omega-6 fatty acids. As a functional marker, it depicts the presence of physiological and nutritional deficiency of essential fatty acids. Conversely, mead acid can replace other PUFAs in the tissues and blood and can lead to platelet hyperactivity, vasoconstriction and altered cell to cell adhesion. Mead acid replacement consequentially lead to the production of pro-inflammatory marker lipoxygenase [29]. In the assessment of EFA status, the use of whole blood plasma may not be a preferred option because fatty acid profiles yields four different classes of fatty acids with different functions. Erythrocyte (RBC) fatty acid provides a more reliable parameter because it reflects bone marrow fatty acid content and plasma RBC phospholipid exchange processes from the preceding 2 – 3 months. In addition, RBC fatty acid that is derived from RBC plasma membrane has a full range of long chain polyunsaturated fatty acid and this gives an indication of the dietary consumption of fatty acids. On the other hand erythrocyte RBCs can be affected by age distribution[29]. 13 Erythrocyte RBCs also reflect the fatty acid composition in the brain in humans: erythrocyte DHA is correlated with brain cortex DHA in a human study and this can indicate that erythrocyte DHA may be a valid marker of brain DHA in humans[30]. An earlier study in rats also demonstrates that red blood cell and neural membranes show similar composition of very-long-chain polyunsaturated fatty acids after dietary modification[31]. A transporter at the blood brain barrier, Mfsd2a, specifically transport DHA in the form of lysophosphatydyl choline-DHA from peripheral tissues into brain tissues[32], and the presence of DHA in the brain tissues enhances cognition. In using the erythrocyte for the assessment of EFAD, there are some ratios that have been established that gives an indication of EFAD. Among these are the triene-tetraene (T/T) ratio, omega-3/omega-6 ratio, palmitoleic acid/linoleic acid ratio, and DHA/AA ratio. Among these ratios the gold standard and the most widely accepted to indicate EFAD is the T/T ratio[29]. Research gap A recommendation ranging from 25-40% energy as total fat intake has been established [6]. However most people in developing countries have lower intakes of fats[33]. Specifically, Ghanaian diets are mainly carbohydrates with inadequate levels of proteins and fats, making the population susceptible to EFAD[34]. FA status is difficult to assess in many developing countries and few studies have been conducted on FA status and children. Recent studies in Ghana utilized FA based supplements[35, 36] but did not assess FA status in Ghanaian children, so this presents a very important research gap. 14 The prevalence of stunting in Ghana is 19% for children below 5 years of age. Although, there have been a reduction in stunting in the past decade, the situation still persist in Northern Ghana, where 33% of all children under 5 years are stunted[37]. Stunting is an indicator of chronic malnutrition which can be caused by inadequate dietary intake. Most of the interventions have focused on vitamin and mineral supplementation and fortification [3]in order to curb stunting, but the situation still persist. ALA and LA and their derivatives can be found in seeds, nuts, oils and animal sources such as fish and meat[33, 38], and some of these sources are uncommon in the diets of some Ghanaians[34] due to cost and this may be increasing the risk of EFAD in Ghana. The quest to find cheap alternative sources of EFA in Ghanaian seeds, nuts and oils is also a gap in literature. In this research, we assessed the FA levels in children 2 to 6 years of age and related them to growth and cognitive function. This is important because there have not been any study in Ghana that has assessed the FA levels of children in this age group. Further, this age group was considered because, at this age, there is rapid growth and development and this is a critical period where interventions are likely to be effective. In addition, after two years of age, most children are weaned off breastmilk and they start consuming family foods which are usually low in fat intake, hence low FA status are expected for this age group. 15 CHAPTER 3: ASSOCIATION OF WHOLE BLOOD N-6 FATTY ACIDS WITH STUNTING IN 2-TO-6-YEAR-OLD NORTHERN GHANAIAN CHILDREN: A CROSS- SECTIONAL STUDY Data in this chapter is published in Adjepong M, et al., (2018) Association of whole blood n-6 fatty acids with stunting in 2-to-6-year-old Northern Ghanaian children: A cross-sectional study. PLoS ONE 13(3) Abstract In Northern Ghana, 33% of children are stunted due to economic disparities. Dietary fatty acids (FA) are critical for growth, but whether blood FA levels are adequate in Ghanaian children is unknown. The objective of this study was to determine the association between whole blood FAs and growth parameters in Northern Ghanaian children 2-6 years of age. A drop of blood was collected on an antioxidant treated card and analyzed for FA composition. Weight and height were measured and z-scores were calculated. Relationships between FAs and growth parameters were analyzed by Spearman correlations, linear regressions, and factor analysis. Of the 307 children who participated, 29.7% were stunted and 8.0% were essential FA deficient (triene/tetraene ratio>0.02). Essential FA did not differ between stunted and non-stunted children and were not associated with height-for-age z-score (HAZ) or weight-for-age z-score (WAZ). In hemoglobin adjusted regression models, both HAZ and WAZ were positively associated with arachidonic acid (p≤0.01), dihomo-gamma-linolenic acid (DGLA, p≤0.05), docosatetraenoic acid (p≤0.01) and the ratio of DGLA/linoleic acid (p≤0.01). These data add to the growing body of evidence indicating n-6 FAs are critical in childhood linear 16 growth. Our findings provide new insights into the health status of an understudied Northern Ghanaian population. Introduction Childhood growth stunting, deemed stunting, is a major nutritional challenge that affects over 165 million children globally [39]. Continents with the highest prevalence of stunting include Africa, Asia, and South America, where stunted and underweight children have a threefold higher risk of mortality [40] compared to well-nourished children. On the African continent, the Sub-Saharan region is most affected by stunting [39]. Observational studies in African countries such as Ghana report roughly 20% of children are stunted, with regional differences in Northern Ghana experiencing stunting rates over 30% [20]. Several fatty acid (FA) supplementation trials in Ghana reported increases in hemoglobin (Hb) levels of pregnant women and may support growth spurts in children [20, 41], however, there is a dearth of research investigating the relationship between circulating FA levels and growth in Ghana. FAs are important for childhood growth and development, and have roles in energy utilization, brain myelination [35], hormone synthesis [41], and FAs serve as substrates for signaling molecules. FAs are synthesized de novo, except for the two essential FAs (EFAs): linoleic acid (LA, an omega-6 [n-6]) and alpha-linolenic acid (ALA, an omega-3 [n-3]) [35, 42]. Humans lack the delta-12 and delta-15 desaturase enzymes required to synthesize LA and ALA, thus, dietary intake is the primary source of EFAs. LA and ALA are elongated and desaturated to produce long chain polyunsaturated fatty acids (LC-PUFAs), and LC-PUFAs serve as substrates for eicosanoids such as prostaglandins, which are involved in cell 17 differentiation and growth [43], and signal transduction [44]. The desaturation of LA and ALA is mediated by the delta-5 and delta-6 desaturase enzymes [45], while several elongases are involved in converting LA and ALA to LC-PUFAs. The n-3 FAs and n-6 FAs have high affinity as substrates for elongase and desaturase enzymes but in their absence, omega-9 (n-9) FAs can serve as substrates for conversion to certain LC-PUFA species, notably the n-9 Mead acid [46, 47]. When an individual's diet is deficient of EFAs, oleic acid (C18:1n9), a non-essential n-9 FA, is converted to Mead acid (C20:3n9) [48]. Mead acid is then incorporated into phospholipids, cholesterol esters, triglycerides, and non-esterified free FAs [47, 49]. It is well accepted that those with EFA deficiency (EFAD), have higher levels of Mead acid and an elevated triene to tetraene ratio (T/T ratio), and the T/T ratio is defined as the ratio of Mead acid (3 double bonds, triene) to arachidonic acid (AA; 4 double bonds, tetraene) [50-53]. EFAD is defined by a T/T ratio > 0.02 in plasma samples [54, 55], and also established when Mead acid [55] levels are above 0.4% in red blood cells (RBCs) [53] and 0.21% in plasma [54]. Northern Ghanaian diets consist mainly of cereals and fruits, with intake of fats and proteins below adequate levels [34]. Additionally, only 36% of Ghanaian children between 6–36 month of age have fats added to their complementary food [37] . These low intakes of dietary fat may increase EFAD in infants and young children. In the Ghanaian diet, sources of LA include peanut and soybean oil, which are also high in ALA [56]. Fish, eggs, poultry, and whole grains are also good sources of EFAs [18, 19, 54]. In Ghana, malnutrition in infants and children below 5 years is mostly caused by inadequate complementary feeding practices[37] which can lead to insufficient EFA consumption . 18 Lipid-based supplementation has shown to increase linear growth in Ghanaian children[57]. However, blood assessments of FA levels are underreported in Ghanaian children and, consequently, EFAD prevalence is poorly characterized. Given the importance of FAs in growth and development, and the high prevalence of stunting in Northern Ghana, the objectives of this study were to assess blood FA levels in 2-to-6- year-old Northern Ghanaian children and their associations with growth, and to characterize blood FA profiles for this population to aid future research and intervention studies. It was hypothesized that whole blood EFA levels in Ghanaian children would positively correlate with the growth measures weight-for-age (WAZ), weight-for-height (WHZ), height-for-age (HAZ) and BMI-for-age z-scores (BAZ). Subjects and methods Study setting The study was conducted in the Northern region of Ghana in the Savelugu-Nanton district [58]. The district covers 2022.6 sq. km with a population density of 68.9 persons per sq. km. The population of Savelugu-Nanton is 139,283 persons with 14,669 households. The average rainfall in the Savelugu-Nanton district is 600 mm. The district is also characterized by high temperatures with an average temperature of 34°C. The district is situated in the Savanna woodland that is capable of sustaining livestock, farming and the cultivation of crops such as rice, groundnuts, yams, cassava, maize, cowpea and sorghum. Over 80% of inhabitants are farmers. The main sources of water in the district are boreholes, rivers and streams, public taps, and pipe borne water. Though the primary source of water for taps and pipe borne water is the same, access to either categorizes 19 households under different income levels, perhaps reflecting differences in hygiene and even nutritional status. Thatch is the main roofing material for housing (50.9%). Illiteracy level is high with 69% of all inhabitants 11 years and above having no education. Some common diseases in the district include malaria, gastroenteritis, upper respiratory tract infection, diarrhea, anemia, and pneumonia. There are three operational community health post zones that deliver health services to the people[59]. The Savelugu-Nanton district was chosen for study as stunting levels are above the national average in the rural communities in this district [60] with overall district stunting level of 38.8% [61]. Additionally, it is one of the few areas in Northern Ghana with road access to rural communities. Ethical approval This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects/patients were approved by the Institutional Review Board at Michigan State University (IRB # 16-557) and the Committee on Human Research Publication and Ethics, School of Medical Sciences, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana (CHRPE/AP/236/16). The parent or caregiver of the participating child gave consent prior to the child’s participation. A script of the written consent was read and translated in Dagbani to the parents or caregivers of the children. The parents or caregivers thumb printed the consent document to give consent. They were assured that participation was voluntary and confidential, and that their information would remain anonymous. 20 Sample size and subjects Children (n=307) between 2 to 6 years of age residing in 5 communities in the Savelugu- Nanton district were recruited for the study. The communities were Janjorikukuo, Pong Tamale, Kparigilanyi, Morglaa and Fazhini. A power analysis was conducted from the results of an earlier study that measured maternal and infant erythrocyte fatty acid intake [38], and the fatty acid variation reported was utilized to run an a-priori sample size calculation for multiple regression based on an estimated medium effect size of 0.5 and significance level p=0.05. This indicated that 242 participants would yield statistical power of 80% [62]. 307 children were enrolled, raising the power to 90%. The exclusion criteria included sick and hospitalized children as well as children who were legally declared intellectually disabled. Data were collected in July 2016. Anthropometric measurements Height of all participants was measured to the nearest 0.1cm with a stadiometer (Seca, USA). Weight was measured using a digital bathroom scale to the nearest 0.1kg (Camry, model number: EB9003, China). All measurements were repeated and averages were reported. The date of birth was recorded from the child’s health card or birth certificate. The sex of the child was also recorded. Blood fatty acid assessment Blood spots (40ul) were collected on a dried blood spot card (DBS) as previously described by Jumbe et al., 2016[63, 64]. A sterile single-use lancet was used in puncturing the tip of the middle finger to obtain drops of blood. The first drop of blood was wiped with 21 a sterilized dry pad. The drops of blood were then collected onto the DBS cards. The cards were stored in a dry, cool environment and shipped to the USA for FA analysis at OmegaQuant Analytics, LLC (Sioux Falls, SD). The average time between sample collection and arrival in the US lab was 8 days. Upon arrival in the US lab, the samples were stored at –80°C for 5 days and then analyzed as previously described [38, 63, 65]. Briefly, a punch from the DBS card was combined with the derivatizing reagent [boron trifluoride in methanol (14%), toluene, and methanol (35:30:35 parts)], shaken and heated at 100°C for 45 minutes. Forty parts of both hexane and distilled water were added after the mixture had cooled. After vortexing briefly, the samples were spun to separate layers and an aliquot of the hexane layer that contained the FA methyl esters was extracted. FA analysis was performed as previously described [66-68]. Unless otherwise stated, whole blood FA proportions are expressed as a percent of total identified FAs. Hemoglobin and malaria status Additional drops of blood from the same puncture site were used to assess hemoglobin concentration using a HemoCue photometer (HemoCue 301, Angelholm, Sweden), and malaria status using an antigen-based malaria rapid diagnostic test kit (Standard diagnostic Inc., Korea). Data reduction and statistical analyses Z-scores were calculated for the growth parameters HAZ, WAZ, and WHZ using WHO Anthro v3.2.2 igrowup package for R [69], to calculate z-scores for children < 5 years of 22 age, and WHO Anthro Plus [70] for children  5. Means and standard deviations were calculated for descriptive analysis. Stunting percentages were calculated based on the WHO standard population and definitions of moderate and severe stunting, wasting, and underweight [71]. FAs were expressed as percent composition of total blood FAs. Mean and standard deviations were calculated for blood FA composition. Total n-3 FA proportions were calculated as ∑ [ALA+ eicosapentaenoic acid (EPA) + docosapentaenoic n-3 (DPA n-3) + docosahexaenoic acid (DHA)]; total n-6 FA proportions were calculated as ∑ [LA + linoelaidic + eicosadienoic (EDA) + dihomo- gamma-linolenic (DGLA) + AA + docosatetraenoic (DTA) + docosapentaenoic n-6 (DPA n-6)]; total n-9 FA proportions were calculated as ∑ [oleic + elaidic + eicosenoic + Mead + nervonic]; total saturated FA proportions were calculated as ∑ [myristic + palmitic + stearic + arachidic + behenic + lignoceric]; total monounsaturated FA (MUFA) proportions were calculated as ∑ [palmitoleic + oleic + palmitelaidic + nervonic + elaidic + eicosenoic]; total polyunsaturated FA (PUFA) proportions were calculated as ∑ [total n-3 + total n-6]. T/T ratio was calculated from the ratio of Mead acid and AA [69]. Product-to-precursor ratios were calculated to estimate PUFA metabolism [72] as follows: EDA/LA to estimate elongase activity, GLA/LA and AA/DGLA to estimate desaturase activity, and DGLA/LA to estimate combined elongase and desaturase activity. All statistical analyses were conducted using software R (R version 3.4.0). Correlations between participant characteristics, anthropometric measurements, and blood FAs were assessed using spearman correlations and graphically displayed using the R package corrplot [73]. Normal probability plots were assessed to verify the validity of regressions. Regression formulas consisted of either the dependent variable HAZ or WAZ, and models 23 were adjusted for each FA and Hb levels (i.e., HAZ = FA + Hb or WAZ = FA + Hb). Hemoglobin was selected as a covariate since it was significantly associated with HAZ and WAZ (p ≤0.01). Regression models were adjusted for Hb and not adjusted for sex as there were few significantly different fatty acids (FAs) between sexes and regression values were unaffected when evaluated with sex adjustment. P-values were considered significant if p≤0.05. Exploratory factor analysis was carried out using the psych package [74]. Briefly, scree plot was used to determine four factors [29]. Palmitelaidic, linoelaidic, and elaidic acids were omitted from the analysis as they were not highly correlated with any other FAs (r<0.3). Varimax rotation was used for orthogonal transformation of the factor loading matrix. FAs correlated with factors r0.5 were considered strongly correlated with the factor, regardless of sign. Factor loading scores were generated for each child and used to calculate regressions for each factor. The regressions were HAZ or WAZ = Hb + Factor. Results Subject characteristics Demographic information is presented in Table 1. In this study, the median age of all 307 children was 46.8 months, with the youngest and oldest being 24.0 and 70.8 months, respectively. There were more males (52.1%) than females (47.9%) in the study. The median height of participants was 96.1 cm, and the median weight of participants was 13.5 kg. Hb levels ranged from 8.4 g/dl to 13.6 g/dl, malaria positivity was detected in 2.9% of the children, and all children were breastfed as infants with 94% of them having been breastfed until 20 months of age (data not shown). The median HAZ, WAZ, WHZ 24 and BAZ were -1.31, –1.16, -0.45, and –0.33, respectively. None of the participant characteristics differed by sex (Table 1). The standard deviations of the HAZ, WAZ, and WHZ distributions were relatively constant and close to the expected value of 1.0 (range: 0.78 – 1.14, data not shown). According to the WHO criteria [74], 29.3% of the children were stunted and 15% were underweight (Table 2). Approximately, 3.5% were categorized as wasting, or had a low BMI for their age. Table 1: Sex differences are not associated with characteristics of participants p-valuea Median (Q1, Q3) Male n=160 Female n=147 Overall n=307 Age (mo) Height (cm) Weight (kg) HAZ BAZ WAZ WHZ Hb (g/dL) 46.8 (37.2, 57.0) 96.1 (80.0, 102) 13.5 (12.2, 15.5) -1.30 (-2.10, -0.69) -0.33 (-0.91, 0.11) -1.20 (-1.70, -0.52) -0.45 (-1.00, 0.02) 11.0 (10.3, 11.7) 48.0 (37.2, 57.6) 96.6 (90.7, 103) 13.9 (12.6, 15.6) -1.32 (-2.10, -0.52) -0.24 (-0.96, 0.11) -1.20 (-1.70, -0.44) -0.39 (-1.10, 0.02) 11.0 (10.3, 11.7) 45.6 (37.2, 56.4) 95.6 (89.1, 101) 13.3 (11.9, 15.3) -1.30 (-2.10, -0.85) -0.37 (-0.80, 0.09) -1.10 (-1.70, -0.58) -0.52 (-0.91, -0.05) 10.9 (10.4, 11.7) 0.73 0.13 0.07 0.50 0.72 0.83 0.40 0.60 The WHO definitions of moderate and severe stunting, wasting, underweight and malnutrition were applied to the data[69]. HAZ, height-for-age z-score; BAZ, BMI-for-age z-score; WAZ, weight-for-age z-score; WHZ, weight-for- height z-score; Hb, hemoglobin. aWilcoxon-Mann-Whitney test was conducted to assess sex differences, p-values presented. Table 2: Nutrition and growth status of the children Severe (<-3SD) Based on Moderate (≤-2SD) Unaffected 70.3% HAZ BAZ WAZ WHZ Stunting 6.07% 23.6% 96.8% Malnutrition 85.0% Underweight Wasting 96.5% The WHO definitions of moderate and severe stunting, wasting, underweight and malnutrition were applied to the data [69], SD, standard deviation; HAZ, height-for-age z-score; BAZ, BMI-for-age z-score; WAZ, weight-for-age z-score; WHZ, weight-for-height z-score. 0.00% 1.60% 0.00% 3.19% 13.4% 3.50% 25 Fatty acid levels in whole blood The median, first, and third quartiles comparing selected FAs of stunted (HAZ≤-2) and non-stunted children are shown in Table 3, and values for all FAs analyzed in our study are presented in Table S1. Approximately 8% of all children in the study had a whole blood T/T ratio greater than 0.02 and 6.8% of the study population had whole blood Mead acid levels above 0.21%. Oleic acid and total n-9 were significantly higher in stunted children, while DHA, the omega-3 index, and total n-3, and AA and DTA were all higher in non-stunted children. There was no significant difference for the T/T ratio or Mead acid between boys and girls. 26 Table 3: Median (Q1,Q3) of selected fatty acid proportions in whole blooda Class Fatty acid Overall Stunted Non-stunted p- valueb ≤0.05 0.64 0.61 0.12 0.32 ≤0.05 0.75 0.89 0.72 ≤0.01 ≤0.05 ≤0.01 20.6 (19.5, 22.2) 0.17 (0.14, 0.23) 0.31 (0.27, 0.40) 0.13 (0.11, 0.17) 0.73 (0.60, 0.90) 21.7 (20.5, 23.4) 21.1 (20.1, 23.7) 0.17 (0.14, 0.25) 0.31 (0.27, 0.36) 0.12 (0.10, 0.15) 0.71 (0.57, 0.92) 22.2 (21.3, 24.5) 20.8 (19.5, 22.6) 0.17 (0.14, 0.23) 0.31 (0.27, 0.39) 0.13 (0.10, 0.16) 0.72 (0.59, 0.91) 21.9 (20.7, 23.6) 0.15 (0.11, 0.21) 0.18 (0.13, 0.24) 0.56 (0.47, 0.67) 2.60 (2.24, 3.03) 3.51 (3.12, 4.00) 2.74 (2.41, 3.20) 0.16 (0.11, 0.24) 0.18 (0.13, 0.24) 0.54 (0.46, 0.67) 2.42 (2.09, 2.76) 3.30 (2.98, 3.78) 2.58 (2.29, 3.03) 0.16 (0.11, 0.21) 0.18 (0.13, 0.24) 0.55 (0.47, 0.67) 2.53 (2.18, 2.96) 3.47 (3.08, 3.95) 2.70 (2.36, 3.17) Oleic Elaidic Eicosenoic Mead Nervonic Total n-9c ALA EPA DPA n-3 DHA Total n-3d O3I LA GLA EDA DGLA AA DTA DPA n-6 Total n-6e GLA/LA EDA/LA DGLA/LA AA/DGLA n-9 n-3 n-6 Ratios aValues represent blood fatty acid (FA) % composition. Stunted defined by height-for-age z-score (HAZ) ≤-2. n-9, omega-9; ALA, alpha-linolenic acid; EPA, eicosapentaenoic acid; DPA n-3, omega-3 docosapentaenoic acid; DHA, docosahexaenoic acid; LA, linoleic acid; GLA, gamma-linolenic acid; EDA, eicosadienoic acid; DGLA, dihomo- gamma-linolenic acid; AA, arachidonic acid; DTA, docosatetraenoic acid; DPA n-6, omega-6 docosapentaenoic acid; n-6, omega-6. bP-value from Wilcoxon-Mann-Whitney test comparing stunted and non-stunted children. cTotal n-9 includes oleic, elaidic, eicosenoic, Mead, and Nervonic. dTotal n-3 includes ALA, EPA, DPA n-3, and DHA. eTotal n-6 includes LA, linoelaidic, GLA, EDA, DGLA, AA, DTA, and DPA n-6. 20.7 (19.4, 21.7) 0.15 (0.12, 0.20) 0.29 (0.24, 0.33) 1.36 (1.18, 1.52) 11.0 (9.94, 11.9) 1.68 (1.44, 1.92) 0.58 (0.47, 0.69) 36.0 (34.4, 37.4) 20.8 (19.7, 22.1) 0.16 (0.12, 0.19) 0.28 (0.24, 0.34) 1.35 (1.18, 1.49) 10.8 (9.60, 11.4) 1.62 (1.36, 1.80) 0.54 (0.46, 0.68) 35.8 (34.5, 36.7) 20.6 (19.2, 21.5) 0.15 (0.11, 0.20) 0.29 (0.25, 0.33) 1.37 (1.19, 1.54) 11.2 (10.0, 11.9) 1.72 (1.48, 1.95) 0.59 (0.48, 0.70) 36.1 (34.3, 37.6) 0.14 0.74 0.41 0.32 ≤0.01 ≤0.01 0.06 0.10 0.93 0.15 0.11 0.19 0.01 (0.01, 0.01) 0.01 (0.01, 0.02) 0.07 (0.06, 0.08) 8.05 (7.21, 8.99) 0.01 (0.01, 0.01) 0.01 (0.01, 0.02) 0.07 (0.06, 0.07) 7.88 (7.04, 8.79) 0.01 (0.01, 0.01) 0.01 (0.01, 0.02) 0.07 (0.06, 0.08) 8.11 (7.31, 9.08) Correlations between fatty acids and growth parameters Spearman correlations were calculated for participant characteristics, anthropometric measurements, and selected FAs (Fig. 2). HAZ was positively correlated with AA (p≤0.01) and DTA (p≤0.01). HAZ was negatively correlated with total n-9 FAs (p≤0.05). WAZ was positively correlated with AA (p≤0.05) and DTA (p≤0.05). Interestingly, height and weight 27 were positively correlated with the ratio of DGLA/LA (p≤0.05), but were not significantly associated with either GLA/LA or EDA/LA. No significant associations were observed between the blood FA levels and BAZ or WHZ. Figure 2: Spearman correlations between participant characteristics, anthropometric measurements, and selected blood FA levels. aSpearman correlation matrix displays r correlation coefficients represented as circles, where large circles represent values closer toward 1 or -1 and smaller circles represent values closer toward 0.1 and -0.1. Blue shades denote positive r correlation coefficients, while red shades denote negative r correlation coefficients. Only results with p0.05 are displayed, thus, empty boxes were not significant. FA, fatty acid; Hb, hemoglobin; WAZ, weight-for-age z-score; WHZ, weight-for-height z-score; BAZ, BMI-for-age z-score; HAZ, height-for-age z-score; n-9, omega-9; ALA, alpha-linolenic acid; EPA, eicosapentaenoic acid; DPA n- 3, omega-3 docosapentaenoic acid; DHA, docosahexaenoic acid; n-3, omega-3; O3I, omega-3 index; LA, linoleic acid; GLA, gamma-linolenic acid; EDA, eicosadienoic acid; DGLA, dihomo-gamma-linolenic acid; AA, arachidonic acid; DTA, docosatetraenoic acid; DPA n-6, omega-6 docosapentaenoic acid; n-6, omega- 6. Total n-9 includes oleic, elaidic, eicosenoic, Mead, and Nervonic. Total n-3 includes ALA, EPA, DPA n- 3, and DHA. Total n-6 includes LA, linoelaidic, GLA, EDA, DGLA, AA, DTA, and DPA n-6. 28 Relationships or associations between fatty acids and growth parameters Table 4 shows the results of the regression analysis for HAZ and WAZ by selected FAs. Since the results of models adjusted for Hb and Hb + Sex were similar, only Hb adjusted models are presented. There was a significant, positive relationship between HAZ and AA (p≤0.01), DGLA (p≤0.05), DTA (p≤0.01), and DGLA/LA (p≤0.01). In addition, WAZ was positively associated with stearic acid (p≤0.05), AA (p≤0.01), DGLA (p≤0.05), DTA (p≤0.01), and DGLA/LA (p≤0.01). Mead acid was positively associated with both HAZ and WAZ. The n-3 PUFAs were not associated with any of the growth parameters. No significant associations were observed between any of the FAs and WHZ or BAZ. A table of all analyzed FA regressions can be found in Supplementary Materials (Table S3) 29 Table 4: Regression results between HAZ, WAZ and selected fatty acids Class Fatty acid Beta ± SE Beta ± SE p-value HAZ WAZ p-value 0.10 0.94 0.56 ≤0.05 0.65 0.10 0.81 0.45 ≤0.01 0.68 0.23 0.77 0.93 0.43 0.62 0.47 ALA EPA DPA n-3 DHA Total n-3c O3I -0.67 ± 0.56 0.07 ± 0.25 0.03 ± 0.37 0.08 ± 0.10 0.04 ± 0.07 0.06 ± 0.08 -0.32 ± 0.43 0.17 ± 0.19 0.20 ± 0.28 0.09 ± 0.08 0.06 ± 0.06 0.08 ± 0.06 -0.04 ± 0.02 -0.02 ± 0.33 -0.35 ± 0.60 2.48 ± 1.21 0.13 ± 0.29 -0.04 ± 0.02 -0.03 ± 0.02 -0.04 ± 0.25 -0.01 ± 0.46 2.08 ± 0.93 0.14 ± 0.23 -0.03 ± 0.02 Oleic Elaidic Eicosenoic Mead Nervonic Total n-9b 0.14 n-9 0.87 0.98 ≤0.05 0.53 0.15 0.45 n-3 0.36 0.48 0.26 0.27 0.23 0.11 n-6 0.78 0.27 ≤0.05 ≤0.01 ≤0.01 0.54 0.26 0.97 Ratios 0.64 ≤0.01 0.92 aModels were adjusted for Hb. Beta ± standard error (SE) presented. HAZ, height-for-age z-score; WAZ, weight-for- age z-score; n-9, omega-9; n-3, omega-3; ALA, alpha-linolenic acid; EPA, eicosapentaenoic acid; DPA n-3, omega-3 docosapentaenoic acid; DHA, docosahexaenoic acid; O3I, omega-3 index; n-6, omega-6; LA, linoleic acid; GLA, gamma-linolenic acid; EDA, eicosadienoic acid; DGLA, dihomo-gamma-linolenic acid; AA, arachidonic acid; DTA, docosatetraenoic acid; DPA n-6, omega-6 docosapentaenoic acid. bTotal n-9 includes oleic, elaidic, eicosenoic, Mead, and Nervonic acid. cTotal n-3 includes ALA, EPA, DPA n-3, and DHA. dTotal n-6 includes LA, linoelaidic, GLA, EDA, DGLA, AA, DTA, and DPA n-6. LA GLA EDA DGLA AA DTA DPA n-6 Total n-6d GLA/LA EDA/LA DGLA/LA AA/DGLA -0.04 ± 0.03 -0.21 ± 0.73 -0.80 ± 0.72 0.39 ± 0.19 0.08 ± 0.03 0.39 ± 0.13 0.18 ± 0.29 0.02 ± 0.02 -0.06 ± 0.03 -0.19 ± 0.95 -1.48 ± 0.94 0.51 ± 0.25 0.12 ± 0.04 0.50 ± 0.17 0.56 ± 0.38 0.03 ± 0.03 0.07 0.84 0.12 ≤0.05 ≤0.01 ≤0.01 0.14 0.18 0.54 ± 13.6 -6.52 ± 14.0 8.34 ± 3.38 0.00 ± 0.03 4.19 ± 17.7 -13.8 ± 18.3 11.7 ± 4.39 0.02 ± 0.04 30 Factor analysis Exploratory factor analysis was conducted to identify FA patterns that may be related to stunting. Scree Plot analysis indicated that four factors should be extracted. Factor loadings (Table 5) show the correlation between individual FAs and the respective factor. Individual factor loadings for each factor were regressed against HAZ and WAZ (Table 6). Factor 1 was significantly associated with HAZ and WAZ (p≤0.01). Several n-6 fatty acids, including AA, DGLA, and DTA were highly correlated with this factor. Factors 2-4 were not significantly associated with either HAZ or WAZ. 31 Table 5: HAZ and WAZ regressed on calculated factors Fatty acid Factor 2 Factor 1 Factor 3 Factor 4 AA DTA DPA n-6 Stearic DGLA 0.89 0.81 0.72 0.64 0.59 -0.09 0.07 0.27 -0.06 0.31 0.08 0.13 0.09 0.24 -0.15 0.14 -0.15 -0.11 0.07 0.20 -0.55 -0.69 -0.23 -0.38 0.05 0.18 -0.19 -0.04 -0.05 0.35 0.49 0.44 0.12 0.23 -0.22 -0.35 0.17 0.04 Palmitic -0.49 Oleic 0.12 Palmitoleic 0.09 Myristic -0.06 GLA -0.16 Mead -0.02 LA -0.26 Arachidic 0.18 Eicosenoic -0.01 Behenic 0.08 Lignoceric 0.21 Nervonic 0.87 DPA n-3 0.82 DHA 0.81 EPA -0.25 ALA EDA 0.08 aVarimax rotated factor-loading matrix generated using the R-package psych. Factors named based on majority of highly correlated FAs. Numbers displayed represent each FA correlation with its respective factor. Correlations ≥0.50 are bolded. AA, arachidonic acid; DTA, docosatetraenoic acid; DPA n-6, omega-6 docosapentaenoic acid; DGLA, dihomo-gamma-linolenic acid; GLA, gamma-linolenic acid; LA, linoleic acid; DPA n-3, omega-3 docosapentaenoic acid; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; ALA, alpha-linolenic acid; EDA, eicosadienoic acid. -0.25 -0.08 0.10 0.04 -0.19 -0.04 -0.20 0.78 0.74 0.73 0.54 0.50 0.06 -0.01 0.08 -0.23 0.47 0.27 0.22 0.80 0.66 0.65 0.62 0.62 -0.17 0.13 -0.29 -0.25 -0.05 0.12 -0.07 -0.06 -0.25 0.15 Table 6: Factor analysis of fatty acidsa HAZ Factor 1 Factor 2 Factor 3 Factor 4 Beta 0.18 0.03 -0.09 0.01 p-value ≤0.01 0.60 0.18 0.86 Beta 0.13 0.01 -0.01 0.02 WAZ p-value ≤0.01 0.82 0.91 0.65 Model: HAZ or WAZ=Factor+Hb aHAZ, height for age z-score; WAZ, weight for age z-score; Hb, hemoglobin. 32 Discussion The purpose of this study was to characterize the whole blood FA levels of Northern Ghanaian children and determine FA associations with growth parameters. Based on Mead acid levels and the T/T ratio, we found that EFAD was low in this population, however, the prevalence of stunting was high (29%). Furthermore, 6.7% of all participants had Mead acid levels above 0.21%, and 8.0% had a T/T ratio greater than 0.02, both lower than previously reported in Tanzania [63]. Interestingly, Mead acid was positively associated with both HAZ and WAZ despite low EFAD. In regression analyses, n-6 LC- PUFAs were inversely associated with stunting, and T/T ratio and total n-3 FAs were not significantly associated with any of the growth parameters. Taken together, our results indicate that whole blood n-6 LC-PUFAs levels are inversely associated with growth stunting, and even though EFAD was low, Mead acid levels were positively associated with growth parameters in this population of Northern Ghanaian children. The levels of EFAs (i.e., ALA and LA) did not differ between stunted and non-stunted children, and were not associated with measures of linear growth. The n-3 LC-PUFAs were not associated with growth in regression analyses, although DHA, total n-3, and the omega-3 index were significantly lower in stunted children compared to non-stunted children. Although whole blood LA levels did not differ, n-6 LC-PUFAs such as AA, DTA, and DGLA were low in stunted children. As DGLA, AA, DTA, and total n-6 FA levels increased in the blood, there was an increase in HAZ, thus, a decrease in stunting. Overall these findings are consistent with previous reports in Tanzanian children [63] and support evidence that n-6 FAs are important in linear growth [75]. 33 Dietary intake of PUFAs is reflective of whole blood PUFA levels [76], therefore, increasing dietary intake of PUFAs in these children may increase whole blood PUFA levels, improve growth parameters, and reduce stunting. AA status is positively correlated with first year growth in preterm infants [77], and AA can regulate gene expression related to cytokine production and osteoclast differentiation [76]. Dietary AA also positively correlates with plasma insulin-like growth factor-1 concentrations [76] which increases hypertrophic cell size and linear growth [78], and is a main predictor of height velocity in children. The mechanisms by which n-6 FAs affect growth and development include serving as substrates for the synthesis of ligands in signal transduction pathways[8]. For instance, AA-derived eicosanoids have key roles in normal growth and development, with prostaglandin E2 functioning in hormone regulation of bone development [75, 76, 79, 80]. Since the n-6 PUFAs DGLA, AA, and DTA were significantly associated with growth parameters, product-to-precursor ratios were assessed to investigate PUFA metabolism. As mentioned LA levels did not differ in regression or dichotomous analyses. The conversion of LA (C18:2 n-6) to DGLA (C20:3 n-6) occurs by: 1) LA elongation to form EDA, then EDA desaturation to form DGLA; or 2) LA desaturation to form GLA, then GLA is elongation to form DGLA [81]. The ratios of GLA/LA or EDA/LA were not significantly associated with growth parameters, however, the ratio of DGLA/LA was significantly associated with growth parameters. The de novo conversion of LA to DGLA utilizes the elongation of very long chain fatty acid protein 5 (ELOVL-5) and delta-6-desaturase enzyme, while the conversion of DGLA (C20:3) to AA (C20:4) utilizes the delta-5- desaturase enzyme. The ratio of AA/DGLA was not significantly associated with growth parameters. It is possible in this population of Northern Ghanaian children that PUFA 34 metabolism is altered through several enzymes involved in FA metabolism. What remains unclear is whether these observations are due to altered PUFA metabolism, lower dietary intakes of PUFA, or both. Gene-diet interactions are known to potentially modulate enzyme activities of some desaturases that are involved in FA metabolism [82]. Fatty acid desaturase-1 (FADS1) and FADS2 genes that encode delta-5-desaturase and delta-6-desaturase, respectively [83], have been hypothesized to be under selective pressure by extreme PUFA diets. The hallmark paper by Fumagalli et al. (2015) reported that isolated populations with extreme diets in varying PUFA composition coped by physiologically adapting gene variants in FADS enzymes [84]. More interestingly, is these FADS gene variants were found to also significantly influence height and weight. We report Mead acid, a PUFA well recognized to accumulate under conditions of EFAD, was positively associated with HAZ and WAZ. This result was unexpected. Mead acid did not differ between stunted and non-stunted children, nor did LA or T/T ratio. One possible explanation is this rural population in Northern Ghana, over the millennia, may have physiologically adapted to EFAD and diets low in PUFAs, similar to the findings of Fumagalli et al. As previously mentioned we report that EFAD was low in this population, despite a prevalence of stunting at 29%. When individuals are deficient in dietary LA, oleic acid is desaturated by delta-6-desaturase, elongated by ELOVL5, and finally desaturated by delta-5-desaturase to form Mead acid [63]. We report the ratio of DGLA/LA was significantly associated with both HAZ and WAZ, and the de novo conversion of DGLA/LA is also dependent on delta-6-desaturase and ELOVL-5. Furthermore, in these children, Mead acid and the ratio of DGLA/LA were highly correlated (Fig. 2). Future research should investigate FADS gene variants in these 35 children to determine if our findings are due to altered metabolism at the biochemical level or due to altered dietary intake. The purpose of this study was to assess blood FA levels in 2-to-6-year-old Northern Ghanaian children and associations with growth parameters. This study cannot be generalized to the entire Ghanaian population, since the study was performed in one village and dietary intake of foods can differ across Ghana. The authors acknowledge the children were not required to fast prior to whole blood collection, blood was collected throughout the day, and these factors may have added variability to our results. However, this variability is expected to be minimal since children in this village consume similar, low-fat meals compared to children in other populations. We did not measure indices of body fat such as mid-upper arm circumference or body composition. In addition, we do not have data on nutrient intake in this population. Nutritional deficiencies aside from EFAD can lead to poor growth in the children. For example, zinc can also affect FA metabolism, and we did not measure zinc. We acknowledge that product-to-precursor ratios provide an indirect estimation of enzyme activity and may not fully reflect biochemical activity. The authors speculate our Mead acid finding may be related to altered PUFA metabolism at the enzymatic level, and conducting this analysis was outside the scope of our current study. Therefore, future researchers should investigate FADS enzyme gene variants in this region of Northern Ghana. To our knowledge this is the first study to assess whole blood FAs in Ghanaian children 2-6 year olds. This study utilized biomarkers of FA status rather than food intake questionnaires to study the associations between FA status and growth. The study was large enough to detect an association between growth metrics and FA levels. Additionally, 36 the use of a validated dried blood spot collection and blood transport system made the study logistically easier to conduct, and the method was also successfully used in a similar study in Tanzania [63]. Our findings add to the growing body of evidence indicating n-6 FAs play a crucial role in linear growth. These data provide new insights into the health of rural Northern Ghanaian children, and provide valuable information for potential intervention studies attempting to combat stunting via nutrient supplementation. 37 CHAPTER 4: WHOLE BLOOD N-3 FATTY ACIDS ARE ASSOCIATED WITH EXECUTIVE FUNCTION IN 2 TO 6-YEAR-OLD NORTHERN GHANAIAN CHILDREN. Data this chapter is published in Journal of Nutritional Biochemistry by Adjepong et al., 2018. Abstract Several studies demonstrate the importance of essential fatty acids (EFAs), and the long chain polyunsaturated FA docosahexaenoic acid (DHA), on cognition and brain development. The objective of this study was to investigate the relationship between whole-blood FAs and executive function in children from Northern Ghana. A total of 307, 2-to-6-year-old children attempted the dimensional change card sort (DCCS) task to assess executive function, and dried blood spot samples were collected and analyzed for FA content. Significant differences in mean % total whole-blood fatty acids were observed between children who could not follow directions on the DCCS test (49.8% of the sample) and those who could (50.2% of the sample). Positive associations with DCCS performance were observed for DHA (β =0.25, p=0.06), total n-3 (β =0.17, p=0.06) and dihomo-gamma-linolenic acid (DGLA; β =0.60, p=0.06). Children with the highest levels of total n-3 and DHA were three and four times, respectively, more likely to pass at least one condition of the DCCS test of executive function than those with the lowest DHA levels. The results of this study indicate an association between n-3 FAs and high-level cognitive processes in children two to six years of age, providing impetus for further studies into possible interventions to improve EFA status of children in developing countries. 38 Introduction Essential fatty acids (EFAs) and their long chain metabolites have crucial roles in human growth, both in fetal and neonatal development [85-87]. They accumulate in the fetus during pregnancy and during early childhood [85]. Long chain polyunsaturated fatty acids (LCPUFA) are also concentrated in the central nervous system [88] playing significant roles in neuronal growth and differentiation of cells and have been associated with cognitive abilities [88-90]. In addition, the brain and retinal function are highly dependent on EFAs, especially for membrane fluidity and signal transduction [26]. Due to these crucial roles of LCPUFAs, poor PUFA status may affect brain development and cognitive abilities in children [90]. There is rapid brain growth in infants and children as evidenced by the 60-fold increase in brain weight from the second trimester to two years of age: 20 to 1200g [85]. Maximal cerebral volume is achieved between 10-15 years of age, but 95% is reached by six years of age [91]. Thus, LCPUFA should be included in the diets of infants and children to ensure optimal brain development [92-94]. EFAs can be found in foods such as peanut and soybean oil, walnuts, fish, eggs, poultry and whole grains ([56, 95], but these are not affordable or available to a large proportion of the population in some developing countries. Specifically, the Ghanaian diet is carbohydrate and protein-rich, but fat-poor,[96] making the population susceptible to EFA deficiency. Several double-blind, randomized control studies in infants and children have established that supplementation with EFA [linoleic acid (LA) , alpha linolenic acid (ALA)] and/or their metabolites [DHA, eicosapentaenoic acid (EPA) and arachidonic acid (AA)] results in improved cognition as evidenced by improved visual acuity and IQ maturation [97], verbal learning, memory[11], progress in myelination, mental and motor 39 development [13] as well as influencing neurological development status[98]. Supplementation results in higher blood levels of EFAs and their metabolites as measured in these studies. Dalton et. al reported that higher supplementation of DHA and EPA-rich fish oil correlated with higher plasma levels of DHA and EPA in 7-9 year olds [11]. This suggests that high levels of circulating blood LCPUFAs can be induced by dietary supplementation and may improve cognitive function in children. Sheppard and Cheatham [99] concluded that LCPUFA influence the cognitive development of children especially with regard to planning and memory processing. A study conducted in Tanzania showed that whole blood FA status was associated with cognitive abilities in children 4-6 year old [64], however no studies of the association between whole blood FAs and cognitive function has yet been conducted in the Ghanaian population to the best of our knowledge. Executive function, which involves inhibition, working memory and task switching [100] is the conscious control of thoughts and actions. It develops in children between the ages of two and ten years [100]. The frontal and temporal lobes of the brain controls executive function [99]. These two regions of the brain contain high amounts of AA and DHA and continue to develop after the second year of life [101]. The dimensional change card sorting (DCCS) task is a validated method commonly used to provide a unitary measure of executive function in young children [100] [102]. In this study, we utilized the DCCS task to assess executive function in Ghanaian children age 2-6. Little is known about cognitive function assessment in the Ghanaian population as well as their association with FAs. In this study, we assess the association between whole blood FA status and executive function in Ghanaian children using a DCCS test of executive function. We 40 hypothesized that whole blood levels of EPA, DHA, and both EFAs (ALA and LA) would be positively associated with performance on the DCCS test. Methods Study site This study was conducted in Savelugu-Nanton; a 2023 sq.km district with a population density of 68.9 per sq. km. in the Northern region of Ghana. The district is situated in the Savanna woodland that is capable of sustaining livestock and many farming practices. The main sources of water in the district are boreholes, rivers and streams, public taps, and pipe borne water. Common diseases in Savelugu-Nanton include malaria, gastro enteritis, respiratory infections, diarrhea, and anemia. The district has three operational community health post (CHP) zones that deliver health services to the people [103]. Subjects and ethical approval This study observed all ethical standards and was approved by the Institutional Review Board at Michigan State University (IRB # 16-557) and the Committee on Human Research Publication and Ethics, School of Medical Sciences, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana (CHRPE/AP/236/16). All the children in the village between 2 and 6 years of age invited to participate, and 313 healthy children were enrolled in this cross-sectional study in July 2016. All participants and their mothers/caregivers verbally consented to participate in the study. 41 Anthropometric measurements Heights of all participants were measured to the nearest 0.1cm with a stadiometer (Seca, USA). Weight was measured using a digital bathroom scale to the nearest 0.1kg (Camry, model number: EB9003, China). The average of two height and weight measurements were recorded. The date of birth was recorded from the child’s health card or birth certificate. The biological sex of the child was also recorded. Height, weight, date of birth and sex data were entered into World Health Organization (WHO) Anthro [104] and WHO AnthroPlus [72] software to calculate height-for-age (HAZ), weight-for-age (WAZ), and BMI-for-age (BAZ) z-scores. Blood fatty acid assessment Blood spots (40ul) were collected on a dried blood spot card (DBS) as previously described in Jumbe et al., 2016 [63]. The tip of the middle finger was punctured with a sterile single-use lancet to obtain drops of blood. A sterilized pad was used to wipe the first drop of blood. The drops of blood were then collected onto the DBS cards. The cards were stored in a dry, cool environment and then shipped to the USA for FA analysis at OmegaQuant Analytics, LLC (Sioux Falls, SD). The average time between sample collection and arrival in the US lab was 8 days. The samples were stored at –20 degrees Celsius and analyzed as previously described [66-68]. Concisely, the DBS card was punched and combined with the derivatizing reagent [boron trifluoride in methanol (14%), toluene, and methanol (35:30:35 parts)], shaken and heated at 100°C for 45 minutes. After cooling, forty parts of both hexane and distilled water were added and briefly vortexed. The samples were spun to separate layers and an aliquot of the hexane layer 42 that contained the FA methyl esters was extracted. FA analysis was performed as previously described [70, 71, 105]. Unless otherwise stated, whole blood FA proportions are expressed as a percent of total identified FAs. Hemoglobin and Malaria status To assess the hemoglobin levels of subjects, additional drops of blood from the same puncture site were used to assess hemoglobin concentration using a HemoCue photometer (HemoCue 301, Angelholm, Sweden). Malaria status was also assessed from a drop of blood using an antigen-based malaria rapid diagnostic test (RDT) kit (Standard Diagnostic Inc., Korea). Cognitive assessment: Dimensional change card sort (DCCS) The DCCS [100, 106] asks that the child sort a series of bivalent cards based on one of two instructed dimension (i.e., either color or shape). Following sorting an initial series of eight cards based upon color, the child is instructed to switch the categorization dimension and sort another series of eight cards based upon shape (see figure 3). This dimensional change in sorting behavior provides an index of executive function as the child must suppress their previously learned set of rules (i.e., sorting by color) and attentional inertia towards those attributes in order to flexibly adjust their behavioral actions and attention to sort the cards by a new set of rules (i.e., sorting by shape) [100, 107]. For each level of the DCCS test, the child was considered to have passed if he/she correctly sorted 6 of the 8 cards in both the pre- and post-switch phases of the task. Given 43 the population of interest and the large developmental spectrum assessed, four levels of the DCCS test were utilized to ensure a robust assessment of executive function. Children who passed the first (instructional) level were allowed to take other 3 levels. Children who failed (scored less than 6 out of 8) the first level were considered to not be able to follow instructions and not allowed to take other levels of the DCCS test. As prior research has demonstrated that children younger than 48 months of age particularly struggle to complete this task, an initial condition was performed to assess if the child’s executive function was sufficiently developed to enable them to follow directions [64, 107, 108]. This condition utilized the same pre- and post-switch procedure as outlined above but utilized monovalent cards that only presented a singular dimension (i.e., either color or shape). If the child was able to pass this initial condition, they were then asked to attempt three additional conditions of the DCCS test. These conditions replicated the traditional DCCS test using bivalent cards, but manipulated the attentional characteristics of the cards by progressively integrating the color and shape attributes to reduce practice effects (Figure 3)[64]. The total number of DCCS test conditions passed was used as an index of executive function[64]. The mother or caregiver was present during all conditions of the DCCS test to observe the process and allow the child to feel comfortable and confident. 44 Figure 3: Illustration of the target cards and test cards used during the pre-switch and post switch Data reduction and statistical analyses A total of 313 children were recruited. A total of 6 children were removed from the analysis due to highly skewed fatty acid profiles failing the Grubs test for a number of fatty acids. The analyses presented herein are from the remaining 307 children. Although the level of fatty acid deficiency in Ghana is unknown, results from previous investigation in another African country [38] was used to calculate an a priori power analysis. Assuming a conservative effect size (f^2 = 0.05), a two-sided alpha of 0.05, and a beta of 0.20 (i.e., 45 80% power) a sample of 242 participants was estimated to provide adequate power. A post hoc analysis of the statistical power using the method of Cohen & Jacob[109] was conducted. Using the data obtained from the analysis of linoleic acid, with 307 subjects and alpha set to 0.05, we had 90% power to detect an R-squared of 0.43. This was adjusted for 3 additional independent variables (Age, Hb and BAZ) with an R-squared of 0.37. Descriptive analyses were conducted to obtain means and standard deviations for all participants. Means between groups (i.e. those who passed at least one DCCS task versus those who did not pass at least one DCCS task) were compared using t-tests (for continuous data) with R software and double-checked with SPSS. Models for linear regression included the FA of interest, and covariates hemoglobin, age and BMI-for-age (BAZ). Hemoglobin concentration was included in our model as a confounder because in similar populations it is a significant predictor of cognitive abilities [110]. Age and BAZ also showed significant association with the dependent variable (total passes). Malaria was not included as a covariate because only 2.93% of the children tested positive to malaria. Also, malaria was not significantly associated with DCCS performance and as such was not included in the model. Factor analysis was conducted using SPSS version 24. Scree plot was used to identify four factors. Trans FAs palmitelaidic, linoelaidic, and elaidic acid were omitted from the analysis as they were not highly correlated with other FAs (r<0.3). Varimax rotation was used for orthogonal transformation of the factor loading matrix. FAs correlated with factors above r=0.5 were considered strongly correlated with the factor, regardless of sign. Factor loading scores generated for each subject were used to calculate regressions for each 46 factor to determine the associations between these factors and performance on the DCCS tasks. The regressions were Total pass = Factor + Age + Hb + BAZ. All statistical analyses were conducted using both SPSS and R software (R version 3.3.0). To determine associations between individual FA groups and executive function, we conducted binary logistic regressions for categorical variables using SAS version 9·4. In all cases, p-value < 0.05 was used to define statistical significance. Results Subject characteristics In this study, 313 children between two and six years of age were enrolled. Six children with outlier myristic acid values were excluded from the study. Demographic information for all 307 participants are shown in table 7. In this study, the average age of all 307 participants was 46.5 months and there were more males (52.1%) in the study than females. The mean height was 96.2 cm and the mean weight was 13.9kg. The mean hemoglobin level of 11.5 was within the normal range [111]. Z-scores were used to calculate the prevalence of stunting (HAZ), malnutrition (BAZ) and wasting (WAZ). According to WHO standards,[112] 70%, 85% and 97% of the 307 participants had normal HAZ, WAZ and BAZ scores respectively. Children who passed the initial condition of the DCCS test were found to be older, taller, and had higher HB levels than children who failed the initial DCCS test (Table 8). Of the 307 children who attempted the DCCS task, 154 children (50.2%) were unable to follow directions as indicated by failing to pass the initial condition of the DCCS, 9 children (2.9%) passed the initial condition but not any 47 other DCCS conditions, 10 children (3.3%) passed two DCCS conditions, 57 children (18.6%) passed three DCCS conditions, and 77 children (25.1%) passed all four DCCS conditions. In comparing the FA levels between both groups, children who passed had significantly lower levels of total saturated fatty acids (p=0.02), but higher omega-6 DGLA (p=0.01) and arachidonic acid (p=0.01), as well as higher n-3 DHA (p=0.05). Table 7: Characteristics of children who attempted the DCCS test Age (mo) Mean 46.5 12.6 SD Sex (male) n % 163 52.1 Range 24.0-70.8 96.2 13.9 Height (cm) Weight (kg) Malaria (%) HB (g/dL) BAZ HAZ WAZ HB, hemoglobin; BAZ, BMI-for-age z-score; HAZ, height-for-age z-score; WAZ, weight-for-age z-score. The WHO definitions of moderate and severe stunting, wasting, underweight and malnutrition were applied to the data [113] 8.40-21.20 72.9-119.9 -4.28-2.43 1.97 11.5 0.17 6.79 -3.40-1.60 0.79 1.13 0.86 8.72 2.41 -2.82-1.83 -0.38 -1.34 -1.12 1.00-2.00 8.40-13.6 48 Pass (n=153) Mean ± SE <0.001 82 50.3% Fail (n=154) P 53.6±0.81 24.0-70.8 39.5±0.88 24.0-69.6 Table 8: Characteristics of children stratified by dimensional change card sort performance for the initial condition (Mean values and standard deviations; numbers and percentages) Anthropometry SFAs Omega-6 FAs Omega-3 FAs HB, hemoglobin; BAZ, BMI-for-age z-score; HAZ, height-for-age z-score; WAZ, weight-for-age z-score. The WHO definitions of moderate and severe stunting, wasting, underweight and malnutrition were applied to the data [69]. 1Total SFA includes myristic, palmitic, arachidic, behenic, lignoceric; 2Total n-6 docosatetraenoic, docosapentaenoic n-6; 3Total n-3 includes alpha linolenic, eicosapentaenoic, docosapentaeonic n-3, and docosahexaenoic; Age (mo) Age range Sex (male) n % Height (cm) Weight (kg) Malaria (%) HB (g/dL) BAZ HAZ WAZ Myristic Palmitic Oleic Total SFA1 LA DGLA AA Total n-62 ALA EPA DHA Total n-33 91.5±0.54 12.8±0.20 1.97±0.01 10.8±0.93 -0.23±0.06 -1.47±0.09 -1.11±0.07 0.22±0.01 21.7±0.15 21.6±0.23 37.7±0.11 20.5±0.16 1.32±0.0.02 10.5±0.13 35.2±0.22 0.18±0.01 0.23±0.02 2.54±0.05 3.54±0.08 100.8±0.62 15.0±0.20 1.97±0.01 11.2±1.07 -0.53±0.06 -1.21±0.09 -1.13±0.07 0.18±0.01 21.1±0.13 21.0±0.21 37.4±0.09 20.7±0.14 1.40±0.0.02 11.0±0.13 36.1±0.18 0.19±0.01 0.22±0.02 2.69±0.05 3.67±0.09 <0.001 <0.001 0.736 0.001 0.001 0.013 0.681 0.011 0.001 0.086 0.018 0.313 0.010 0.007 0.002 0.259 0.776 0.048 0.196 includes linoleic, linolaidic, ϒ-linolenic, eicosadienoic, di-homo-gamma-linolenic, arachidonic, 81 49.7% FA levels in whole blood and regression analysis Whole blood fatty acid levels of the 307 children whose data were analyzed are presented in Table 9. The mean levels of the essential fatty acids ALA and LA were, 0.18 and 20.56, respectively. The mean % of total FA in whole blood for DHA was 2.6, for EPA 0.22 and for the omega-3 index 4.5. Regression analysis between selected FAs and DCCS performance, adjusting for age, BAZ and Hb, is shown in Table 10. The n-3 FA DHA, as well as n-6 FA DGLA were positively associated with DCCS performance. To test the 49 hypothesis that whole blood levels of EPA, DHA, and both EFAs (ALA and LA) would be positively associated with executive control as indexed by performance on the DCCS tasks, multiple linear regression using EPA, DHA, ALA, LA, Hb, age, and BAZ was conducted. The model explained 38% of the variation (r2=0.384; adjusted r2=0.370, p0.001). DHA (β=0.40, p=0.02) and ALA (β=0.14, p=0.01) were significant contributors to the model, both being positively associated with performance on the DCCS test. A full model including all 25 single FAs as well as Hb concentrations, age, and BAZ was significant (p<0.001) and explained about 43% of the variance (r2=0.429; adjusted r2=0.374). In this full model, DHA (β=0.58, p=0.02) and ALA (β=2.75, p=0.007) were positively associated with DCCS performance. However, the effects of independent FAs and the covariates could not be determined due to high levels of collinearity leading to poor tolerance and variance inflation in the model. 50 Table 9: 1Whole blood fatty acid proportions in Ghanaian children (Mean ± SE, n=307). Class Mean ± SE Fatty acid Range SFA n-3 FA n-6 FA n-9 FA Desaturases Myristic Lignoceric Palmitelaidic Palmitic Behenic Arachidic Total SFA2 Alpha-linolenic Eicosapentanoic Docosahexaenoic Omega-3 Index Total n-33 Linoleic Arachidonic acid GLA DGLA Docosatetraenoic Eicosadienoic Total n-64 Mead acid Oleic Eicosenoic Nervonic Total n-95 SCD n-7 SCD n-9 D6d D5d Palmitoleic 0.20±0.01 1.24±0.02 0.03±0.001 21.4±0.10 0.86±0.01 0.36±0.04 37.56±0.07 0.18±0.01 0.22±0.01 2.62±0.04 4.55±0.52 3.61±0.05 20.56±0.11 10.77±0.09 0.17±0.004 1.36±0.01 1.69±0.02 0.29±0.004 35.67±0.14 0.14±0.003 21.3±0.16 0.34±0.01 0.75±0.01 22.4±0.15 0.02±0.001 1.61±0.02 0.067±0.001 8.11±0.08 0.367±0.01 0.02-0.57 0.30-2.23 0.003-0.14 17.0-28.0 0.33-1.44 0.20-0.67 32.8-41.1 0.04-0.78 0.05-3.11 1.35-6.12 2.97-11.74 2.02-11.1 14.9-27.2 5.31-14.8 0.04-0.49 0.69-2.44 0.74-2.68 0.13-0.62 27.2-41.1 0.04-0.39 15.7-31.2 0.17-0.65 0.22-1.34 17.4-31.8 0.003-0.05 1.01-3.40 0.03-0.12 4.39-12.7 0.06-1.40 23.0±0.15 39.3±0.16 0.013±0.00 Total MUFA Total PUFA T/T ratio 1Expressed as FA % proportion (n=307); 2Total SFA includes myristic, palmitic, arachidic, behenic, lignoceric; 3Total n-3 includes alpha linolenic, eicosapentaenoic, docosapentaenoic n-3, and docosahexaenoic; 4Total n-6 docosatetraenoic, docosapentaenoic n-6; 5Total n-9 includes oleic, elaidic, eicosanoic, Mead, nervonic SFA, saturated fatty acid; SCD n-7, stearoyl CoA desaturase n-7; 8SCD n-9, stearoyl CoA desaturase n-9; D6d, delta-9-desaturase; D5d, delta-5-desaturase; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid; T/T, triene-to-tetraene linolaidic, ϒ-linolenic, eicosadienoic, di-homo-gamma-linolenic, arachidonic, 17.7-32.4 29.7-45.0 0.003-0.04 includes linoleic, 51 Regression results for total pass (n=307) B ± SE P Table 10: Regression results for performance on the dimensional change card sort test and selected fatty acids (FA). Class SFA n-3 FA n-6 FA n-9 FA Desaturases Fatty acid Myristic Lignoceric Palmitelaidic Palmitic Behenic Total SFAa Alpha-linolenic Eicosapentanoic Docosahexaenoic Omega-3 Index Total n-3b Linoleic Arachidonic GLA DGLA Docosatetraenoic Total n-6c Mead acid Oleic Eicosenoic Nervonic Total n-9d SCD n-7 SCD n-9 D6D D5D Palmitoleic Total MUFA Total PUFA T/T ratio -0.18 ± 0.72 -0.17±0.26 -5.38±3.13 -0.03±0.05 -0.43±0.44 0.00±0.07 1.04±0.70 0.27±0.32 0.25±0.13 0.17±0.09 0.17±0.09 -0.02±0.04 0.02±0.05 0.94±1.21 0.60±0.32 0.11±0.22 0.01±0.03 -0.81±1.55 -0.02±0.03 0.11±0.79 -0.15±0.37 -0.02±0.03 -0.46±9.39 -0.21±0.23 10.04±5.62 -0.09±0.06 -0.02±0.41 -0.02±0.03 0.02±0.03 -9.76±15.4 0.80 0.50 0.09 0.53 0.33 0.96 0.14 0.40 0.06 0.07 0.07 0.60 0.63 0.44 0.06 0.61 0.84 0.60 0.55 0.89 0.69 0.44 0.96 0.35 0.08 0.10 0.96 0.44 0.43 0.53 Other Model: Total pass= Fatty acid of All Children + Age + BAZ + Hemoglobin) aTotal SFA includes myristic, palmitic, arachidic, behenic, lignoceric. bTotal n-3 includes alpha-linolenic, EPA, DPA n-3, and DHA. cTotal n-6 includes linoleic, linolaidic, GLA, eicosadienoic, DGLA, arachidonic, DTA, DPA n-6. dTotal n-9 includes oleic, elaidic, eicosanoic, Mead, nervonic. SFA, saturated fatty acid; SCD n-7, stearoyl CoA desaturase n-7; SCD n-9, stearoyl CoA desaturase n-9; D6d, delta- 6-desaturase; D5d, delta-5-desaturase. Polytomous logistic regression To determine the association between individual fatty acids and DCCS performance, a binary logistic regression analysis was performed to assess the relationship between the children who passed the initial phase of the DCCS test versus those who failed, for every 52 unit increase in the fatty acid of interest. Since a histogram plot shows a skewed data, all fatty acids of interest were categorized into three groups based on biological significance, instead of tertiles, and results shown in Table 11. Children with highest DHA levels per the category (DHA >4.0%) were four times more likely to pass the initial phase of the DCCS than children with the lowest DHA levels (DHA <2.0%) (OR: 3·5; 95%CI: 1.3·9.2 – 35·3; p=0·02). Table 11: Associations of significant fatty acids, as tertiles, with dimensional change card sort test performance. FAa DHA (%)  2.0 >2.0 to  3.0 > 3.0 omega-3 index  4.0 >4.0 to  5.0 Total n-3 (%) > 5.0  3.0 >3.0 to  4.0 > 4.0 Test for exposure Test for trend OR[95% CI]b 1 1.61 [0.71, 3.69] 3.50 [1.33, 9.24] 1 1.65 [0.82, 3.31] 2.31 [1.01, 5.29] 1 1.08 [0.52, 2.26] 2.91 [1.20, 7.06] OR (p trend)c 1.91 (0.008) 1.52 (0.047) 1.72 (0.02) a Whole blood FAs were separated into tertiles. b Test for exposure was conducted to determine if increases in FA tertiles, compared to the lowest tertile, were associated with passing the DCCS test compared to failing. Odds ratio (OR) [95% Confidence Interval (CI)] are displayed. c Test for trend was conducted to determine if increases in FA tertiles were associated with passing the DCCS test compared to failing. Odds ratio (p-value) are displayed. b,c Polytomous logistic regression was used to regress BMI category on FA tertiles. All data are referenced against children who failed the first DCCS test. Both test for trend and test for exposure were adjusted for age, HB, malaria and BAZ. P-values bolded if p≤0.05 or italicized is p≤0.09 and >0.05. Factor analysis Factor analysis was conducted to identify group effects and to bypass the problems of collinearity. When factor analysis was used to determine how combinations of the FAs might be associated with performance on the DCCS tasks, four main factors emerged. 53 The factor loading matrix is shown in Table 12. Multiple linear regression using the four factors showed no factor to be significantly associated with performance on the DCCS tasks (Table 13). When combined with Hb concentrations, age, and BAZ, these parameters explained 37% of the variance (r2=0.369; adjusted r2=0.354) in the performance on the DCCS tasks Table 12: Factor analysis of fatty acids.a Fatty acid Factor 1 0.89 AA DTA DPAn6 Stearic DGLA Oleic Palmitic ALA Elaidic Behenic Arachidic Eicosenoic Lignoceric Nervonic Eicosadienoic Linoelaidic Factor 2 Factor 3 Factor 4 0.12 0.16 0.11 0.25 -0.10 -0.12 -0.25 -0.24 -0.03 0.76 0.76 0.70 0.60 0.57 0.46 0.08 -0.07 0.09 0.29 -0.05 0.33 0.21 0.26 -0.25 0.07 -0.28 -0.17 0.11 -0.24 -0.05 0.14 0.02 0.13 -0.15 -0.11 0.06 0.20 -0.47 0.05 -0.26 0.17 -0.02 -0.26 0.19 0.05 0.19 0.08 -0.06 0.80 0.71 0.65 0.58 -0.70 -0.56 -0.34 -0.18 0.31 -0.07 -0.07 0.45 0.39 0.16 0.03 Palmitoleic GLA Myristic Mead LA DPAn-3 DHA EPA Palmitelaidic aVarimax rotated factor-loading matrix generated using the R-package psych. Factors named based on majority of highly correlated FAs. Numbers displayed represent each FA correlation with its respective factor. Correlations >0.40 are bolded. -0.10 -0.17 0.01 0.00 -0.23 0.07 0.02 0.08 -0.10 -0.25 0.03 -0.39 0.15 -0.15 0.13 0.24 -0.21 0.13 0.79 0.66 0.64 0.63 -0.62 0.10 -0.09 -0.09 0.21 0.15 -0.06 0.12 -0.17 -0.02 0.87 0.80 0.80 0.24 54 Table 13: Regression† results for performance on the dimensional change card sort test and fatty acid (FA) factors Factor Parameter estimate Standardized parameter estimate p-value Factor 1 Factor 2 Factor 3 Factor 4 Age HB BMI-for-age (BAZ) *All significant associations (P<0·05). † Model: total passes= factor 1 + factor 2 + factor 3 + factor 4 + Age + Hb concentration +BMI-for-age z-score; model P value, P<0·001; r2 0.369; adjusted r2 0.354. 0.07 0.002 -0.09 0.08 0.08 0.02 -0.22 0.04 0.001 -0.05 0.05 0.56 0.01 -0.10 0.41 0.98 0.27 0.33 <0.01* 0.78 0.04* Discussion Our data generally support the hypothesis that children with higher whole blood levels of EFAs (LA and ALA) as well as DHA and EPA were more likely to pass the DCCS test, an indicator of executive function. The multiple linear regression model using EPA, DHA, ALA, LA, Hb, age, and BAZ was significant and explained 38% of the variation in DCCS performance. However, when utilizing individual regression analysis, the essential fatty acids ALA and LA were not significantly associated with DCCS performance, nor in the factor analysis. Yet, regardless of the type of analysis herein, we show that children with higher whole blood levels of DHA, total n-3 and the omega-3 index calculation were more likely to pass the DCCS test. Consistent with this observation, children with higher blood n-3 FA levels were more likely to pass the DCCS test. When blood levels of DHA and total n-3 FAs increase, children exhibited improved cognition. This is consistent in a number of randomized, controlled, human supplementation studies which have shown that when full term infants were 55 supplemented with LCPUFA that contained DHA, AA and LA, there was an improvement in visual acuity and IQ maturation[97]. Supplementation of children 3 -10 year olds with fish oils that contained DHA, eicosapentaenoic acid (EPA) and gamma linolenic acid (GLA) also resulted in improved non-verbal cognitive development [114]. Also, there was an improvement in short term memory when children 6 to 10 years old were supplemented with ALA, DHA, LA and micronutrients [12]. In addition, maternal supplementation of DHA and EPA was beneficial in cognitive function of their offspring, suggesting that there are long term effects of the supplementation on offspring [115]. Although the amount of DHA in the brain is variable depending on dietary intake, DHA and AA are the most highly concentrated PUFA in neural phospholipids including subcellular membranes [116]. The mechanisms by which DHA and n-3 FA affects cognition include myelination of axons [13]. DHA and n-3 FAs influences membrane fluidity, neurotransmitter receptor activity and nutrition of nerve cells [42, 117], as well as disruption of lipid rafts that affect signal transduction pathways [118]. Specifically, the cell membrane is a phospholipid bilayer, which contains lipid rafts. Lipid rafts are specialized lipid domains that differ in lipid composition by their cholesterol and sphingolipid content. They serve as a center for assembling signaling molecules, and their domains are disrupted under certain stimuli optimizing plasma membrane function and influencing membrane fluidity. DHA and n-3 FA acyl chains can exhibit conformational changes eliciting a stimulus that can disrupt the domains of the lipid raft. When DHA/EPA are incorporated into the rafts, the cholesterol molecules are redistributed to non-rafts leading to declusteration of the raft system. The non-raft proteins are sequestered into declustered rafts and this triggers a 56 downstream signaling activating cell receptor membranes for communication between cells. This communication between cells can enhance cognitive function. Further, the nerve cell membrane determines the amount of nutrient that can pass through the cell [118]. Rigid membranes does not allow adequate nutrients to get into the cells and the arrangement of the cellular domains is dependent of the presence of double bonds [119]. The presence of numerous double bonds in the DHA and n-3 FA molecular structure increases the fluidity of the membrane allowing nutrients to get into the cells [119], hence n-3 FAs help in the nourishment of cells making cells healthy and less prone to injury. Additionally, the nerve cell membrane contains proteins that act as receptors for some neurotransmitters, transmitting signals across a synapse. Also, fluidity of membranes allows receptors to recognize neurotransmitter and sends the message they contain. These factors provide evidence that n-3 FAs and DHA have significant roles in cognitive function. The omega-3 index (EPA+DHA), a measure of n-3 FAs in red blood cells, influences cardiovascular health and has also been associated with cognitive abilities. Together, DHA and EPA are involved in many aspects of brain function including blood-brain barrier integrity and brain blood flow [116]. This study demonstrated that the omega-3 index was positively associated with performance of the DCCS test. The direction of association was consistent with studies conducted by van der Wurff et al, 2015 [120], which reported that there was a strong positive association between omega -3 index and letter digit substitution test, indicating that children with higher omega-3 index may have a faster information processing speed and less impulsivity. In addition, in this study, low n-3 intake was associated with a decrease in DCCS test performance. Consistent with previous 57 studies by Sheppard and Cheatham, 2017 [121], children with higher n-6s and lower n- 3s performed poorly on the DCCS test. Sheppard and Cheatham, 2017 [121] reported that plasma n-6 and n-3 FA levels predicted and affected performance on working memory and planning tasks. In general, the hippocampus and frontal cortex of the brain is responsible for memory and executive function [121].These parts of the brain are sensitive to changes in n-3 FAs because of the role n-3 FAs play in neurotransmitter concentration, receptor density and function and neuronal growth [121]. In addition, a study in a mouse model indicated that supplementation with only DHA, (an n-3 FA) or AA (an n-6 FA) was insufficient for optimal development [122] supporting the essentiality of both n-6 and n-3 FAs in growth and development. Omega-3 and n-6 FA levels do change with age; omega-3 levels increasing over a lifetime while omega-6 levels decrease [123]. While these results reflect changes in FA levels over decades, there was no significant difference in FA levels for different age categories within a small range for our study. In another study comparing FA changes in 3-8-year old European children, Wolters et. al reported increasing n-6 FA, LA, with age and no significant association with n-3 FA DHA [124]. This study has a number of strengths. This study utilized an objective biomarker to assess dietary fatty acid intake other than conventional and less precise methods such as food frequency questionnaire or the diet history techniques. Food frequency questionnaires are not highly accurate at estimating circulating blood levels of LCPUFAs [125]. Also, the study includes a diverse panel of FAs and estimated desaturase activities, first of its kind in the Ghanaian population. In addition, in this study we culturally adapted a standard DCCS test to suit this population and the overall observed performance is consistent with 58 previous investigations conducted in the Tanzania[64], USA [126] and Scotland [107]. This investigation builds off our prior work in the Tanzanian population [64], to demonstrate these relationships despite potential genetic variation in FA biosynthesis across these populations. With regards to limitations, this study was a cross sectional study and all associations that are reported are correlative rather than causative. It was conducted in the Savelugu- Nanton municipality and hence the results cannot be generalized to the entire Ghanaian population or other areas in the world. The collection of blood for this study was done throughout the day with no fasting required and this may increase variability in the whole blood FA measurements. However, variation in this setting are likely to be minimal as children from the village consume relatively similar and low-fat meals compared to children in other settings. Whole blood samples for lipids were analyzed and are not able to differentiate amongst the source compartment of the lipid. Another limitation to this study is that, the children in this population may have other nutritional deficiencies or diseases/infections that may cause poor cognition, however, the study accounted for at least two factors that are known to be linked to cognition: malaria and low hemoglobin levels. Finally, although the authors did not collect socio-economic data from the parents/caregivers in this population, there is likely to be little variability in these factors since the population is relatively homogeneous [103]. In summary, whole blood n-3 FA levels, in particular n-3 fatty acids, are associated with executive function in this cohort of Ghanaian children. Whether n-3 FA supplementation earlier in life would improve cognitive performance in these children would need to be examined in a randomized trial. 59 CHAPTER 5: QUANTIFICATION OF FATTY ACID AND MINERAL LEVELS OF SELECTED SEEDS, NUTS AND OILS IN NORTHERN GHANA Data in this chapter is under review in the Journal of Food Science Technology by Adjepong et al., 2018 Abstract Fatty acids (FAs) and micronutrients are required for growth and development of children. The objective of this study was to determine the mineral and FA composition of Northern Ghanaian foods. Seven seeds and three oils were collected from a local market. Freeze- dried seeds and food grade oils were packaged and shipped to the US. FAs were quantified by GC/MS. ANOVAs were conducted on FA concentrations and Tukey’s post hoc test was used to compare foods. Palm oil contained significantly higher amounts of the saturated fatty acid (SFA) palmitic acid (293.1 mg/g; p<.001). Shea butter (292.0 mg/g) and palm oil (246.5 mg/g) were highest in oleic acid (p<.001). Soybean was significantly higher in alpha-linolenic acid (ALA) (2.98 mg/g; p<.01). Neri (68.4mg/g) and fermented dawadawa (56.3mg/g) had significant amounts of total PUFAs (p<.0001). Iron levels in soybean (353 mg/kg), neri (282 mg/kg) and fermented dawadawa (165 mg/kg) were relatively high. The identification of the nutrient content of these foods can increase its utilization in homes and in industrial settings especially in the formulation of complementary foods and new recipes. Together, these foods may be useful for future intervention to curb stunting and potential cognitive impairment in this population. 60 Introduction Essential fatty acids (EFAs) and minerals are important dietary components required in human growth and development. They are key in cell membrane formation and proper development of brain and nerve cells. FAs also serve as major dietary energy source playing crucial roles in cell differentiation and metabolism [8, 127]. Although FAs are important in the human diet, their intakes in African diets are low [33]. Specifically, the dietary composition of Ghanaians is mainly starchy roots and cereals with levels of protein and fat below adequate levels. Though the dietary supply meets population energy requirements, the low dietary diversity in the population may be causing a deficiency in proteins and fats [18]. Although Ghana has achieved its goal according to the Millennium Development Goal 2 (MDG), uneven wealth distribution and poverty still prevails in the Northern sector of the country. In addition, only 36.6% of Ghanaian children from 6 to 24 months old have fats added to complementary foods [19]. Low dietary diversity in some parts of Northern Ghana, especially in food insecure households, has led to low dietary intake of fats [128] and may contribute to EFA deficiency in the population. The prevalence of mineral deficiencies is high in Ghana as well [129], with iron-deficiency anemia more prevalent in Northern Ghana than other regions [20]. The national stunting levels among Ghanaian children below five years old is 19%. However, in Northern Ghana, it stands at 33% [20]. Whole blood FAs are associated with stunting [63] and cognition [64]; hence, the availability and inclusion of adequate amounts of FAs and minerals in the diet is important. EFAs cannot be synthesized in the body and must be provided by the diet. EFAs are abundant in foods such as soybean oil, canola oil, flaxseed oil, nuts, and animal products. 61 Linoleic acid (LA), an omega-6 (n-6) FA, and alpha-linolenic acid (ALA), an omega-3 (n- 3) FA, are EFAs. Arachidonic acid (AA), is a downstream metabolite derived from LA, while eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are metabolites derived from ALA. EFAs and their long-chain metabolites are important for cognition, growth and immune function [33, 130]. Seafood, especially fish, is a good source of EPA and DHA. However, fish is expensive and not sufficiently available to Ghanaians, evidenced by the fact that the amount of fish required to meet the nutritional needs of Ghanaians exceeds current fish production (marine, inland, and aquaculture) [131]. It has further been reported that the distribution and consumption of fish in Northern Ghana is low [132], leading to a high probability of low EFA intake in the population. The need for the exploration of alternative, available and affordable sources of EFAs is important. Mineral deficiencies are prevalent in developing countries [129, 133, 134]. Although the prevalence of anemia in Ghanaian children 6 to 59 months old is 66%, the Northern region reports 82.1%, with iron deficiency as a major contributor [20]. Mineral deficiencies are also associated with growth stunting and cognitive impairment [135, 136] because some minerals such as iron and zinc are cofactors needed in EFA metabolism. For example, zinc deficiency impairs the conversion of LA to AA and ALA to EPA and DHA [137]. The evaluation of the mineral composition of seeds and nuts has been conducted in some studies, though the available data are highly variant due to environmental factors such as soil type, agronomic practices and climate [138, 139]. Geographical differences can directly affect the concentration of minerals in crop plants, and dietary mineral content, impairing EFA metabolism and potentially leading to growth stunting and cognitive impairment in the population. 62 The West African food composition (WAFC) table is a nutrient database which publishes an estimation of nutrients such as proteins, carbohydrates, total fat, vitamins, and minerals in food. [140] However, it contains many weaknesses: 1) The composition table contains the total fat composition of foods with no information on individual FA contents. 2) There is little information on the origin of the foods analyzed; 3) The data represent average values derived from compositional data of 8 countries, not just Ghana; 4) Most of the mineral and vitamin data in the table are based on information obtained from several non-African countries. Due to these limitations, the nutrient content listed in the WAFC table does not represent the foods in Ghana, and analyses that quantify individual FAs, especially EFAs, and mineral content of local foods in Ghana are needed. This study quantified the amount of EFAs, very long chain PUFAs (VLCPUFAs) and minerals in local foods available at a market in Northern Ghana. Children in this population have high levels of stunting which is associated with a potential EFA deficiency. Therefore, identifying FAs in local foods can be a step to increase utilization of such foods in homes and in industrial settings especially in the formulation of complementary foods and new recipes. Materials and Methods Preparation of selected seeds, oils, and nuts in Ghana Seeds, nuts, and oils (Table 14) were purchased at a local market in Tamale, Northern Ghana. Oils obtained from the markets were transferred to opaque dark plastic containers to prevent lipid oxidation during transport and storage to the US for fatty acid analysis. Local seeds and nuts collected from the market were transported to a lab where they were crushed, freeze-dried and stored in containers for transport to the US for fatty acid 63 analysis to prevent lipid oxidation. All samples were shipped to the Michigan State University Food Science and Human Nutrition laboratory. Once received, the samples were purged with high purity nitrogen and stored at -20°C until analysis. Freeze drying technique was chosen because of shipping duration and power outages in this region. The technique also ensured sample integrity. The extraction yield of freeze-dried samples are comparable to air-dried samples. [141]. In addition, the quantity of LA and ALA from freeze-dried samples is at least as abundant as air-drying methods [142]. Crude seed oil extraction All glassware used was washed with high performance liquid chromatography (HPLC) grade organic solvents to remove mineral residues and FA contaminants. Lipids were extracted from seed material as previously described [143], but modified as specified [144]. In brief, a total of 400 mg freeze-dried seed material was incubated for 2h at RT in 12 mL of a 2:1 (v/v) mixture of HPLC grade chloroform (Avantor Performance Materials, Inc., Center Valley, PA) and methanol containing 100 µg butylated hydroxytoluene (BHT)/mL (Sigma–Aldrich, St. Louis, MO) to release lipids into solution. Extracted seed samples were filtered using lipid-free filters (FGE Healthcare UK Limited, Buckinghamshire, UK) into glass tubes containing 2.5 mL of 0.88% (w/v) aqueous KCl (J.T. Baker, Phillipsburg, NJ) to separate aqueous and organic layers. The tubes were centrifuged and the bottom organic layer was transferred to a new tube and dried under high-purity nitrogen. The total crude seed oil was weighed. 64 Methylation of oils to FAMEs, neutralization, and FAME isolation The samples were resuspended in chloroform/methanol (2:1 v/v, 100 µg BHT/mL) to obtain a final lipid concentration of 20 mg/mL. The resuspended oils were prepared for methylation as described by Cequier-Sanchez et al. [143]. In brief, 100 µL of lipid extract solution was transferred to clean 16×100 mm Teflon-lined screw-capped glass tubes. To each sample, 200 µg of the internal standard, Stearic acid d35 (Sigma - Aldrich, Lot #TP1700V) suspended in HPLC-chloroform, was added. The resultant mixture was dried under high-purity nitrogen at RT. The samples were methylated with 2% acidified methanol as described by Agren et al. [145] and modified by Pickens et al. [70]. The mixture was neutralized and isolated as previously described. [146]. Briefly, the mixture was neutralized with salt buffer and FAMEs were extracted with hexane. FAME identification, analysis, and data Processing Resuspended FAMEs were transferred to GC vials with glass inserts for analysis. Prior to analysis, the sample injection order was randomized. FAMEs were identified and quantified (in triplicate) using a dual stage quadrupole (DSQ)II quadrupole GC/MS (Thermo Scientific, Waltham, MA) equipped with a DB-23, 30-m column (0.25 mm id; Agilent Technologies, Santa Clara, CA) using helium as a carrier gas. GC/MS temperature profile and selective ion monitoring (SIM) were performed as previously described [144]. Identification and quantification of individual FAMEs were done with standard FAME mixture (Part# CRM47885; Lot# LC06601V; Supelco, Bellefonte, PA) and standard curves were prepared as previously described [144]. Detected FAME concentrations below the lower limit of quantification (LLOQ) are defined for each FA in 65 Tables 2-4. DHA, EPA, and linoelaidic acid were below the LLOQ in all samples analyzed and were excluded from the tables. FAME peak integration and quantification was performed using TargetLynx V4.1 (Waters, Milford, MA) based on the FAME standard’s retention time and SIM ions and ratios. The concentrations of re-suspended FAMEs were normalized to the amount of food-grade oil (i.e. palm oil) or crude seed oil and total seed material for the seed samples (i.e. neri seed). Mineral analysis Freeze-dried samples were analyzed for their mineral content. The samples DAW, FDD, NER, SAM, SOY, PNT and BAB were analyzed by a third-party contractor. The minerals that were analyzed include zinc, iron, potassium, phosphorus, sodium, magnesium, manganese and calcium. Identification and quantification of minerals was performed using ICP emission spectroscopy (ICP_S: 28, AOAC International no. 984.27, 985.01, and 2011.14) [144]. For each sample, concentrations of minerals are expressed as mg mineral/kg sample. Statistical Analyses Mean fatty acid concentrations (mg/g) and standard deviation values are presented in Tables 2-4. Parametric one-way ANOVA was conducted for each FA and p-values are given. Concentration values below the LLOQ were excluded from all analyses. Tukey’s honest significant difference (HSD) post-hoc test was used for multiple comparisons of significant models. Statistical analyses were conducted using R (R version 3.3.0). 66 Results FA composition of the seeds, nuts, and food-grade oils are reported as the concentration of each FA in mg per gram of food sample (Table 2–4). Table 14 describes the FAs and oils that were analyzed. Samples were found to contain varying levels of several FAs, as shown in the data. Table 14: Foods analyzed Food Abbreviation Baobab seed Dawadawa BAB DAW Fermented dawadawa FDD Neri seed Soy bean Peanut Sesame seed Shea butter (oil) Palm oil Palm kernel oil NER SOY PNT SAM SHB PAL PKO Saturated fatty acids General Description Adansonia sp. Parkia sp. Parkia sp. Seeds of an ancient tree, whose fruits possess a velvety shell Seeds of a perennial tree, African locust bean plant Fermented seeds of the African locust bean plant Cucumropsis sp. Glycine sp Climbing vine, flattened seeds, seeds were freeze dried A legume, bean seeds Arachis sp. A root of a tropical legume Sesamum sp. Seeds from a domesticated oil seed plant Vitellaria sp. An ivory-coloured fat extracted from African shea nut Elaeis sp Elaeis sp. Crude oil processed locally, red-orange, high in beta carotene Oil derived from the kernel of the oil palm All samples analyzed were found to contain saturated FAs. Food-grade oils were more abundant in saturated fats than seeds (Table 15). SHB, PAL, and PKO were high in total saturated FAs (343.4 mg/g, 330.8 mg/g, and 153.3 mg/g, respectively). PAL was significantly higher in palmitic acid than all other foods (293.1 mg/g; p<.001). SHB was highest in stearic acid content (313.8 mg/g; p<.001). PKO contained the highest amount of myristic acid (82.8 mg/g; p<.001). SAM (27.8 mg/g), SOY (6.36 mg/g), DAW (17.3 mg/g) 67 Table 15: Saturated fatty acids expressed in mg FA/g crude oil or mg FA/g seed (Mean ± SD). Sample ID Baobab seed Dawadawa Fermented dawadawa Neri seed Soy bean Peanut Sesame seed Shea butter Palm oil Palm kernel oil Sample Type Seed Seed Seed Seed Seed Nut Seed Oil Oil Oil Myristic C14:0 Palmitic C16:0 0.15±0.02B 0.02). None of FAs were associated with growth parameters except total saturated fats and linoleic acid which showed trending negative significance. When the data were compared to previously reported data from Northern Ghana, the analysis showed that most n-3 FA levels were significantly higher and n-6 FA levels lower in the Southern Ghana population (p<0.001). Fish and seafood consumption in this Southern cohort is high and could account for low essential FA deficiency and lower stunting rates. Introduction Growth stunting, a condition of impaired development, is a strong indicator of chronic malnutrition and a major global nutritional challenge in Ghana [20]. Growth stunting affects over 165 million children globally, with Africa, Asia and South America reporting a higher prevalence [151]. Several countries in Sub-Saharan Africa are mostly affected 78 by stunting [151]. Studies in Ghana as reported by the Ghana demographic health survey (GDHS) states that 19% of Ghanaian children under five years of age are stunted [20]. Stunting typically becomes permanent once established and may be caused by poor maternal health, frequent childhood infections and inadequate nutrient intake [5]. There have been numerous interventions implemented to curb stunting including dietary supplementation and fortification of vitamins and minerals. Specifically, recent fatty acid (FA) supplementation studies in Ghana demonstrate that, lipid-based supplements increased hemoglobin (Hb) levels of pregnant women and growth spurts in children [35, 41]. However, these studies do not characterize the FA content in whole blood. Further, research investigating the relationship between circulating FA levels and growth is scarce in this population. FAs have numerous physiological functions in human growth and development. For example, polyunsaturated FA (PUFA)-derived eicosanoids can activate transcriptional factors to influence stem cell proliferation and differentiation [8], and essential FAs (EFAs) help build structural barriers to prevent energy loss by the accumulation of LA into the stratum corneum [82]. These FAs can be metabolized into molecules with specificity for receptors whose signal transduction pathway results in changes to linear growth [82]. EFAs are those FAs which cannot be produced in the body because the human body lacks specific enzymes required for their de novo biosynthesis. Linoleic acid (C18:2n6, LA) and alpha-linolenic acid (C18:3n3, ALA) are the two main EFAs in the human diet [43, 44]. These omega-6 (n-6) and n-3 FAs are substrates for the desaturases (delta-5 and delta-6 desaturase) and elongases that produce the long chain (LC) PUFAs such as docosahexaenoic acid (DHA), arachidonic acid (AA) among others [48]. The elongase 79 enzymes usually prefer n-6 and n-3 FAs as substrates, but in their absence, they convert n-9 FAs into some LC-PUFA such as Mead acid, the appearance of which is a hallmark of EFA deficiency [47, 49]. More specifically, when an individual's diet is deficient of EFAs, oleic acid (C18:1n9), a non-essential n-9 FA, is converted to Mead acid (C20:3n9)[50]. Mead acid is then incorporated into phospholipids, cholesterol esters, triglycerides and non-esterified free FAs [51, 52]. Therefore, in EFA deficiency (EFAD), there are elevated levels of Mead acid with a decrease in the production of other EFA metabolites such as arachidonic acid. The ratio of Mead acid (3 double bonds, triene) and AA (4 double bonds, tetraene), termed as triene to tetraene ratio (T/T ratio), is a functional biomarker for EFAD [53-55]. EFAD is defined by a T/T ratio > 0.02 in plasma samples [53, 54], and also established when Mead acid [47] levels are above 0.4% in red blood cells (RBCs) [23] and 0.21% in plasma [54]. Child undernutrition is prevalent in Ghana due to low dietary diversity and poor infant and young child feeding (IYCF). Specifically, only 15% of breastfed Ghanaian children met minimum standards of IYCF practices with respect to both dietary diversity and feeding frequency[20]. Additionally, Ghanaian children 6-23 months are also affected by poor feeding practices such as inadequate complementary feeding. Sometimes, the complementary foods are deficient in essential nutrients such as proteins and fats [58]. Further, anemia, which is often described as an indicator of poor nutrition and poor health, is also prevalent in many parts of Ghana, with all the regions having a prevalence greater than the 40% World Health Organization (WHO) cut off [19]. These dietary intake patterns coupled with poor infant feeding practices could increase EFAD in infants and young children. Apart from fish, eggs, poultry and whole grains are also good sources of EFAs 80 [150, 152]. The dietary sources of the parent EFAs in Ghanaian diets are peanut (both ALA and LA-rich), melon seeds (LA-rich) and soy bean (ALA-rich) [56]. Insufficient consumption of foods which are good sources of EFAs may be leading to growth impairment. Recently, some small-scale studies in Ghana using lipid-based supplementation have showed an increase in linear growth in Ghanaian children [35, 41], however, blood assessment of FA levels are typically not reported in Ghanaian children and the prevalence of EFAD has been poorly documented. Considering the significance of FAs in growth and development, the objective of this study was to assess blood FA levels in 2-to-6-year-old Ghanaian children and their association with growth. Materials and Methods Study setting The study was conducted in the Upper-Manya Krobo district whose district capital is Asesewa located in eastern region of Ghana. The district covers 859.1 sq. km with a population of 72,092. The district comprises of 13,111 households with an average household size of 4.6 persons per household. The rainfall ranges from 900mm to 1500mm with temperature ranging from 26°C to 32°C. The district lies within the semi- deciduous forest and savanna zone. Palm, dawadawa, mango, neem and acacia interspersed with shrubs are the major vegetation in the district. Agriculture, forestry and fishing constitute the largest industry in the locality employing over 72% of the workforce aged 15 years and above. Boreholes, tubewell, pumps, rivers and streams constitute the sources of water supply of household for domestic purposes. Metal sheet is the main 81 roofing material for housing (87.9%). Illiteracy level is high with 33.3% of all inhabitants 11 years and above having no education. Like other communities in Ghana, the inhabitants are at risk of diseases and other contagious illnesses. The community has one hospital, 3 maternity homes, 4 health centers and 15 Community Health Posts[153] Sample size and subjects Children (n=209) between 2 to 6 years of age residing in communities in the Upper Manya Krobo district, Ghana were recruited for the study. A sub sample from a larger cohort, recruited as previously described [154] were enrolled in this study. Communities that were inaccessible for more than two weeks during a given period were also excluded. At household level, a household with a target child who had a medical/birth defect that affect eating and normal growth was excluded (eg. cerebral palsy). Data were collected from March to July 2017. An a-priori power analysis was conducted from the results of an earlier study that measured maternal and infant erythrocyte FA levels [38]. Ethical Standards Disclosure This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects/patients were approved by the Ethics review board at McGill University Canada(IRB#180-1013), the Institutional Review Board at Michigan State University (IRB # 16-557) and the Nogouchi Memorial Institute for Medical Research ethics committee (IRB # 027/13-14). Written informed consent was obtained from all subjects/patients. A script of the written consent was read and translated in Twi, Krobo and Ewe to the parents or caregivers of the children. The parent or caregiver 82 of the participating child gave consent prior to the child’s participation. The parents or caregivers thumb printed the consent document to give consent. They were assured that participation was voluntary and confidential, and that their information would remain anonymous. Anthropometric measurements Heights of all participants were measured to the nearest 0.1cm with a Shorrboard stadiometer (Weigh and Measure LLC, USA). Weight was measured using a digital bathroom scale to the nearest 0.1kg (Tanita BMB-800, Japan). The average of two height and weight measurements were recorded. The date of birth and gender was recorded from the child’s health card or birth certificate. Height, weight, date of birth and sex data were entered into World Health Organization (WHO) Anthro [69] and WHO AnthroPlus [72] software to calculate height-for-age (HAZ), weight-for-age (WAZ), weight-for-height (WHZ), and BMI-for-age (BAZ) z-scores. Blood fatty acid assessment Capillary blood sample (40ul) was obtained by puncturing the middle finger using a sterile single-use lancet as previously described by Jumbe et al. [63, 65]. The first drop of blood was wiped with a sterilized dry pad. The drops of blood were then collected onto the dried blood spot cards, which are pre-treated with anti-oxidant cocktail. The cards were stored in a dry, cool environment and shipped to the USA for FA analysis at OmegaQuant Analytics, LLC (Sioux Falls, SD). On average, the time from sample collection time to 83 arrival to the US was 8 days. The samples were stored at –80°C till they were analyzed as previously described [66-68]. Briefly, the cards were punched and combined with the derivatizing reagent [boron trifluoride in methanol (14%), toluene, and methanol (35:30:35 parts)]. The mixture was shaken and heated at 100°C for 45 minutes. Forty parts of both hexane and distilled water were added after the mixture had cooled. The mixture was vortexed and then separated into distinct layers. An aliquot of the hexane layer that contained the FA methyl esters was extracted. FA analysis was performed as previously described [70, 71, 105]. Unless otherwise stated, whole blood FA proportions are expressed as a percent of total identified FAs. Hemoglobin and malaria status A hemocue photometer (HemoCue 301, Angelholm, Sweden) was used to determine the hemoglobin concentration. The malaria status was determined using an antigen-based malaria rapid diagnostic test (RDT) kit (Standard diagnostic Inc., Korea). These tests were conducted using additional drops of blood from the same punch. Dietary intake assessment Using a structured questionnaire, the intake of various foods consumed by subjects within 24 hours prior to blood sample collection was assessed. The foods that were of interest include fish and seafood, diary, meats and fruits. 84 Data reduction and statistical analyses Z-scores for the growth parameters HAZ, WAZ, BAZ and WHZ were calculated using WHO Anthro [69]. Means and standard deviations were calculated for descriptive analysis. Based on the WHO standard population and definitions of moderate and severe stunting, wasting, and underweight percentages were calculated [74]. The FA values presented here are expressed as percent composition of total blood FAs. Total n-3 FA proportions were calculated as ∑ [alpha-linolenic + eicosapentaenoic acid (EPA) + docosapentaenoic n-3 + docosahexaenoic acid (DHA)]; total n-6 FA proportions were calculated as ∑ [linoleic + linoelaidic + eicosadienoic + dihomo-gamma-linolenic + arachidonic + docosatetraenoic + docosapentaenoic n-6]; total n-9 FA proportions were calculated as ∑ [oleic + elaidic + eicosenoic + Mead + nervonic]; total saturated FA proportions were calculated as ∑ [myristic + palmitic + stearic + arachidic + behenic + lignoceric]; total MUFA proportions were calculated as ∑ [palmitoleic + oleic + palmitelaidic + nervonic + elaidic + eicosenoic]. T/T ratio was calculated from the ratio of Mead acid and AA[29]. FA product/precursor ratio was used to estimate the desaturase activity [155] as follows: D5D= AA/DGLA; D6D= DGLA/LA. FA composition of children who were stunted and those who were not, as well as FA levels of northern vs southern Ghana children, were compared using two sample t-test. Normal probability plots were assessed to verify the validity of regressions. The regression model for HAZ, BAZ, WHZ or WAZ was adjusted for Hb and malaria. Hemoglobin was used as a covariate as it was significantly associated with HAZ and WAZ (p ≤.01). Malaria was also adjusted in the model as a significant percentage (11%) of the children tested positive to RDT/malaria test. Regression formulas consisted of either the 85 dependent variable HAZ, WAZ, WHZ, or BAZ, and models were adjusted for each FA and Hb levels (e.g., HAZ = FA + Hb + malaria). Regression models were not adjusted for sex as there were few significantly different FAs between sexes and regression values were unaffected when evaluated with sex adjustment. P-values were considered significant if p≤ 0.05. All statistical analyses were conducted using software R (R version 3.3.0) and verified with SPSS version 24 (IBM). Regression models were not adjusted for sex as there were few significantly different fatty acids (FAs) between sexes and regression values were unaffected when evaluated with sex adjustment. P-values were considered significant if p≤ 0.05. All statistical analyses were conducted using software R (R version 3.3.0) and verified with SPSS version 24 (IBM). Results Subject characteristics The demographic information for the subjects are presented in Table 19. The mean age of all 209 subjects was 38.31±9.98 months. There were more males (51.7%) than females (48.3%) in the study. The average height of participants was 91.45±7.10 cm, and the average weight of participants was 12.78±2.13 kg. The average hemoglobin level was 10.90g/dl. A positive malaria test was detected in 11.00% of the children. The mean HAZ, WAZ, WHZ and BAZ were -1.35, -1.04, -0.41 and -0.27 respectively. The standard deviations of the HAZ, WAZ, and WHZ distributions were relatively constant and close to the expected value of 1.0 (range: 0.91 – 1.01). Using the WHO guidelines and criteria[74], 86 22% of the children were stunted, 12.9% were underweight and 3.4% were wasted. (Table 20). Table 19:Demographic characteristics of participants Age (mo) Height (cm) Weight (kg) HAZ BAZ 38.31±9.98 91.45±7.10 12.78±2.13 -1.35±0.91 -0.27±1.01 38.46±9.95 91.92±7.04 13.08±1.99 -1.40±0.88 -0.15±0.88 Overall (n=209) (n=108) Mean ± SD Male WAZ WHZ Hb (g/dL) -1.04±0.97 -0.41±1.01 10.90±1.36 -1.00±0.86 -0.32±0.89 10.8±1.45 Female (n=101) 38.15±10.06 90.96±7.16 12.47±2.24 -1.28±0.94 -0.39±1.12 -1.08±1.08 -0.49±1.12 11.00±1.25 HAZ, height-for-age z-score; BAZ, BMI-for-age z-score; WAZ, weight-for-age z-score; WHZ, weight-for-height z score; Hb, hemoglobin.The WHO definitions of moderate and severe stunting, wasting, underweight and malnutrition were applied to the data[69]. Fatty acid levels in whole blood The mean fatty acid levels for overall, stunted and non-stunted children in Southern Ghana are shown in Table 21. Approximately 10.5% of all children in the study had whole blood T/T ratio greater than 0.02 and 4.5% had Mead acid levels above 0.21%. For all 209 children enrolled in the study, the average percent whole blood measured was 40.2% saturated FAs, 23.3% total monounsaturated FAs, 29.1% total omega-6 and 7.2% total omega-3 FAs. The major component components of omega-6 and omega-3 FAs were LA Table 20: Nutrition and growth status of children. Severe (<-3SD) Based on Moderate (≤-2SD) Unaffected Stunting Malnutrition Underweight Wasting HAZ, height-for-age z-score; BAZ, BMI-for-age z-score; WAZ, weight-for-age z-score; WHZ, weight-for-height z- score; aThe WHO definitions of moderate and severe stunting, wasting, underweight and malnutrition were applied to the data[69]. 17.70% 1.90% 10.50% 2.40% 4.30% 1.00% 2.40% 1.00% 78.00% 97.10% 87.10% 96.60% HAZ BAZ WAZ WHZ 87 and DHA, accounting for 16.7% and 5.1% of whole blood respectively. Overall, the mean blood FAs % of total did not differ between stunted and non-stunted children. Of interest, there was no significant difference between Mead Acid when comparing stunted vs non- stunted children. Although not significant, children who were stunted had lower blood levels of DHA (4.95%) than those who were not stunted (5.13%) or the overall group (5.09%). Children who were stunted has significantly higher ratios of delta-5 (p=0.054) and delta-6 (p=0.027) desaturase activity. However, no significant difference was found between T/T ratio of both groups. 88 Table 21: Mean fatty acid levels for overall, stunted and non-stunted children in Southern Ghana (Values expressed as mean ± standard deviation) Class Not Stunted (n=163) Saturated FAs MUFAs n-3FAs n-6 FAs Ratios Fatty Acid Myristic Palmitic Stearic Arachidic Behenic Lignoceric Total sat Palmitoleic Oleic Acid Eicosenoic Nervonic Total MUFA ALA EPA DPA DHA Total n-3 Omega-3 Index LA GLA EDA Eicosatrienoic AA DTA DPA Total n-6 Stunted (n=46) Overall (n=209) 0.82±0.59 0.76±0.41 25.7±1.62 25.8±1.53 11.9±1.16 12.1±1.14 9.18±1.56 0.31±0.06 0.84±0.41 0.59±0.16 0.90±0.34 0.95±0.39 40.2±1.54 40.5±1.46 0.84±0.41 0.88±0.50 21.0±2.68 20.8±2.56 0.34±0.14 0.37±0.13 0.83±0.30 0.87±0.30 23.3±2.70 23.2±2.62 0.25±0.10 0.27±0.12 0.80±0.35 0.74±0.23 1.01±0.21 1.04±0.21 5.09±0.98 4.95±0.95 7.15±1.34 8.03±1.37 7.00±1.19 7.80±1.23 16.7±1.92 16.6±2.43 0.21±0.09 0.20±0.10 0.03±0.05 0.25±0.06 1.24±0.23 1.30±0.25 9.18±1.56 9.09±1.29 1.03±0.25 1.09±0.22 0.45±0.13 1.02±0.26 29.1±2.60 29.0±2.71 0.84±0.63 25.6±1.65 11.9±1.16 0.31±0.06 0.56±0.15 0.88±0.32 40.1±1.56 0.82±0.39 21.0±2.72 0.33±0.14 0.82±0.29 23.3±2.73 0.25±0.09 0.81±0.38 1.00±0.21 5.13±0.98 7.19±1.37 8.09±1.40 16.7±1.76 0.21±0.08 0.25±0.05 1.22±0.22 9.21±1.64 1.02±0.26 0.44±0.13 29.1±2.57 0.09±0.06 4.18±0.08 0.07±0.02 7.72±1.72 7.72±1.72 P Value 0.46 0.592 0.209 0.752 0.215 0.251 0.108 0.411 0.596 0.156 0.248 0.818 0.645 0.22 0.364 0.263 0.379 0.204 0.589 0.61 0.156 0.048 0.645 0.118 0.262 0.779 0.920 0.561 0.027 0.054 0.054 0.09±0.01 0.09±0.07 Mead acid n-6/ n-3 ratio 4.19±0.81 4.26±0.84 Delta-6-Desaturase 0.08±0.02 0.08±0.02 Delta-5-Desaturase 7.61±1.68 7.22±1.48 T/T ratio 7.61±1.68 7.22±1.48 aValues represent blood fatty acid (FA) % composition. Stunted defined by height-for-age z-score (HAZ)≤-2. n-9, omega-9; ALA, alpha-linolenic acid; EPA, eicosapentaenoic acid; DPA n-3, omega-3 docosapentaenoic acid; DHA, docosahexaenoic acid; LA, linoleic acid; GLA, gamma-linolenic acid; EDA, eicosadienoic acid; DGLA, dihomo-gamma-linolenic acid; AA, arachidonic acid; DTA, docosatetraenoic acid; DPA n-6, omega-6 docosapentaenoic acid; n-6, omega-6. bP-value from Wilcoxon-Mann-Whitney test comparing stunted and non-stunted children. cTotal n-9 includes oleic, elaidic, eicosenoic, Mead, and Nervonic. dTotal n-3 includes ALA, EPA, DPA n-3, and DHA. eTotal n-6 includes LA, linoelaidic, GLA, EDA, DGLA, AA, DTA, and DPA n-6. 89 Table 22: Regression results between HAZ, WAZ, and selected fatty acids WAZ Saturated fats Fatty acid Myristic Acid -0.10±0.12 -0.09±0.11 p-value B±SE HAZ B±SE 0.38 MUFAs n-3FAs n-6FAs Palmitic Acid Stearic Acid Total sat Palmitoleic Acid Oleic Acid Eicosenoic Acid Nervonic Acid Total MUFA ALA EPA DPA DHA Total n-3 LA GLA EDA Eicosatrienoic AA DTA DPA Total n6 -0.04±0.04 -0.04±0.06 -0.07±0.04 -0.08±0.15 -0.00±0.02 -0.78±0.46 0.04±0.22 -0.01±0.02 -0.57±0.66 0.00±0.18 -0.20±0.31 0.06±0.07 0.02±0.05 0.04±0.03 0.77±0.68 0.50±1.25 0.05±0.28 0.01±0.04 -0.27±0.25 -0.13±0.48 0.03±0.03 0.38 0.50 0.08 0.61 0.92 0.89 0.86 0.80 0.39 1.00 0.51 0.39 0.65 0.22 0.25 0.69 0.87 0.75 0.29 0.78 0.32 0.00±0.04 -0.01±0.06 -0.03±0.05 0.01±0.16 0.01±0.03 -0.48±0.49 0.14±0.23 -0.01±0.03 0.32±0.70 0.04±0.19 0.20±0.33 0.03±0.07 0.02±0.05 -0.02±0.04 0.82±0.72 0.75±1.33 0.31±0.30 -0.01±0.05 0.02±0.27 0.40±0.51 -0.01±0.03 p-value 0.39 0.83 0.92 0.56 0.96 0.81 0.32 0.53 0.78 0.65 0.85 0.53 0.72 0.68 0.68 0.26 0.57 0.30 0.90 0.95 0.43 0.83 0.77 0.48 0.55 0.02±0.08 0.03±0.05 5.51±9.21 n-6/n-3 ratio Omega-3 Index T/T ratio Ratios Model: HAZ=fatty acid + hemoglobin+malaria; WAZ= fatty acid + hemoglobin+malaria). aValues represent blood fatty acid (FA) % composition. Stunted defined by height-for-age z-score (HAZ)≤-2. n-9, omega-9; ALA, alpha-linolenic acid; EPA, eicosapentaenoic acid; DPA n-3, omega-3 docosapentaenoic acid; DHA, docosahexaenoic acid; LA, linoleic acid; GLA, gamma-linolenic acid; EDA, eicosadienoic acid; DGLA, dihomo-gamma-linolenic acid; AA, arachidonic acid; DTA, docosatetraenoic acid; DPA n-6, omega-6 docosapentaenoic acid; n-6, omega-6. bP-value from Wilcoxon-Mann-Whitney test comparing stunted and non-stunted children. cTotal n-9 includes oleic, elaidic, eicosenoic, Mead, and Nervonic. dTotal n-3 includes ALA, EPA, DPA n-3, and DHA. eTotal n-6 includes LA, linoelaidic, GLA, EDA, DGLA, AA, DTA, and DPA n-6. -0.07±0.08 0.02±0.05 -4.53±9.83 0.42 0.48 0.65 Relationships or associations between fatty acids and growth parameters Table 22 shows the results of the regression analysis between HAZ and WAZ for selected FAs. There was no significant association between FAs and WAZ and HAZ except for total saturated fats (p=0.08) which was associated with HAZ. Table 23 shows a regression 90 between WHZ and BAZ and some selected fatty acids. With the exception of LA which was related with WHZ (p=0.08) and BAZ (p=0.07) all other FAs were not associated with any of the growth parameters. 91 Table 23: Regression results between WHZ, BAZ, and selected fatty acids (model: WHZ=fatty acid + hemoglobin + malaria; BAZ= fatty acid + hemoglobin + malaria) WHZ BAZ Saturated fats MUFA n-3 FAs n-6 FAs Fatty acid Myristic Palmitic Stearic Total sat Palmitoleic Oleic Eicosenoic Nervonic Total MUFA ALA EPA DPA DHA Total n-3 LA GLA Eicosadienoic B±SE -0.10±0.12 0.02±0.05 0.03±0.06 0.03±0.05 0.12± 0.17 0.02±0.03 0.02±0.51 0.19±0.24 0.01±0.03 0.88±0.73 0.03±0.20 0.46 ±0.34 -0.02±0.07 0.01±0.05 -0.07±0.04 0.78±0.75 0.50±1.39 Eicosatrienoic Acid 0.52±0.39 AA DTA DPA Total n6 -0.03±0.05 0.27±0.28 0.66±0.53 -0.04±0.03 Ratios n-6/Total n-3 ratio -0.12±0.09 Omega-3 Index -0.01±0.05 T/T ratio -8.27±10.2 p-value B±SE p-value 0.41 0.65 0.59 0.5 0.50 0.56 0.97 0.43 0.43 0.23 0.87 0.18 0.80 0.89 0.08 0.30 0.67 0.09 0.56 0.33 0.22 0.19 0.18 0.86 0.42 -0.07±0.12 0.03±0.05 0.03±0.06 0.03±0.05 0.10±0.17 0.02±0.03 0.11±0.51 0.19±0.24 0.02±0.03 1.06±0.73 0.03±0.20 0.49±0.34 -0.03±0.07 0.01±0.05 -0.07±0.04 0.54±0.75 0.51±0.39 0.52±0.31 -0.03±0.05 0.28±0.28 0.72±0.53 -0.04±0.03 -0.12±0.09 -0.01±0.05 11.4±10.2 0.55 0.55 0.63 0.50 0.56 0.57 0.83 0.43 0.44 0.15 0.87 0.15 0.72 0.91 0.07 0.48 0.72 0.10 0.53 0.29 0.16 0.17 0.18 0.81 0.27 aValues represent blood fatty acid (FA) % composition. Stunted defined by height-for-age z-score (HAZ)≤-2. n-9, omega-9; ALA, alpha-linolenic acid; EPA, eicosapentaenoic acid; DPA n-3, omega-3 docosapentaenoic acid; DHA, docosahexaenoic acid; LA, linoleic acid; GLA, gamma-linolenic acid; EDA, eicosadienoic acid; DGLA, dihomo-gamma-linolenic acid; AA, arachidonic acid; DTA, docosatetraenoic acid; DPA n-6, omega-6 docosapentaenoic acid; n-6, omega-6. bP-value from Wilcoxon-Mann-Whitney test comparing stunted and non-stunted children. cTotal n-9 includes oleic, elaidic, eicosenoic, Mead, and Nervonic. dTotal n-3 includes ALA, EPA, DPA n-3, and DHA. eTotal n-6 includes LA, linoelaidic, GLA, EDA, DGLA, AA, DTA, and DPA n-6. 92 Mean fatty acid levels for Northern and Southern Ghana, compared Next, we compared the mean fatty acid levels between our previously published data from Northern Ghanaian children [156] to these data from Southern Ghanaian children. Interestingly, there were highly significant differences in mean whole blood % for most of the fatty acids (p<0.001; Table 23). Specifically, the mean whole blood levels of n-3 FAs such as ALA, EPA, DHA, DPA n-3, and omega–3 index were significantly higher (p<0.001) in the Southern Ghana population than in the Northern Ghana population. Mean level of DHA and the omega-3 index, for example, were 2.62% and 8.03% respectively in the Northern Ghana population compared to 5.09% and 4.55% in the Southern Ghana population. Further, the mean whole blood n-6 FAs LA, AA. DTA, DPA n-6 and were higher in the Northern Ghana population (p<0.001). The T/T ratio and Mead acid were higher in Northern Ghana population than the in Southern Ghana population (p<0.001; table 24). This indicates a higher rate of EFA in the Northern Ghanaian population. 93 Table 24: Mean FA levels for Northern and Southern Ghana, expressed as mean ± standard deviation. Northern Ghana Class Fatty acid (% of total) Southern Ghana N=209 N=307 P-Value 0.25±0.10 0.80±0.35 1.01±0.21 5.09±0.98 7.15±1.34 8.03±1.37 0.18±0.12 0.22±0.26 0.58±0.18 2.62±0.64 3.61±0.91 4.55±0.92 n-3 FAs n-6 FAs ratios 1Total n-3 includes alpha linolenic, eicosapentaenoic, docosapentaenoic n-3, and docosahexaenoic; 2EPA+DHA 3Total n-6 includes linoleic, linolaidic, ϒ-linolenic, eicosadienoic, di-homo-gamma-linolenic, arachidonic, docosatetraenoic, docosapentaenoic n-6; T/T, triene-to-tetraene:ALA, alha linolenic acid; EPA, Eicosapentaenoic acid; DHA, Docosqahexaenoic acid; LA, linoleic acid, GLA, Gamma linoleic acid, ; DTA, ALA EPA DPA DHA Total n-31 Omega-3 Index2 LA GLA Eicosadienoic Eicosatrienoic AA DTA DPA Total n-63 n-6/n-3 ratio TT Ratio Mead acid 20.6±1.85 0.17±0.07 0.29±0.07 1.36±0.26 10.8±1.61 1.69±0.37 0.59±0.17 35.7±2.53 16.7±1.92 0.21±0.09 0.25±0.05 1.24±0.23 9.18±1.56 1.03±0.25 0.45±0.13 29.1±2.60 0.013±0.005 0.14±0.05 0.010±0.007 0.09±0.06 10.3±2.00 4.20±0.81 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 docosatetraenoic acid, docosapentaenoic acid 94 Dietary intake of proteins in Southern Ghana Protein intake in the population is summarized in table 25. The records of protein intake shows that 92.82% consumed fish or sea food 24 hours prior to data collection. The records also show that 23.92%, 18.18% and 17.22% of the children consumed meat from poultry, eggs and dairy products respectively. Discussion The objective of this study was to describe the whole blood FA levels of Southern Ghanaian children and to determine the association between FAs and growth parameters. The prevalence of stunting in in our sample was 22% which is higher than the 17% stunting rate reported by the GDHS in the Eastern region of Ghana. However, the stunting rate in this population is lower than what is reported in the Northern Ghana [157]. The prevalence of EFAD was 4.64% and 10.6% based on the Mead acid levels and T/T ratio respectively. These values are both lower than what was previously reported in Tanzania [63]. With the exception of total saturated FAs and LA which showed a trending significance with some growth parameters, none of FAs were associated with growth parameters. When the blood levels of FAs from this study were compared to previously reported data from Northern Ghana, most n-3 FAs levels were significantly higher in the Southern Ghana population (p<0.001) and most n-6 FAs level were significantly higher in the Northern Ghana population (p<0.001). 95 Mean omega-3 index in the Southern Ghana population is significantly higher (p<0.001) than what was reported in an earlier study in Northern Ghana [156]. Whole blood levels of DHA and EPA are highly dependent on dietary intake. Although, there is conversion of ALA to its metabolites including EPA and DHA, there is limited conversion of ALA to DHA in mammalian species [158]. It is therefore likely that the high EPA and DHA levels were obtained from the diet and contribute to the omega-3 index. While the intake of carbohydrates are similar across populations in Ghana[159], it is likely that location or local culture may contribute to limited access of foods rich in DHA and EPA. Over 90% of the subjects consumed fish and sea food in the 24 hours prior to data collection. This consumption pattern is higher when compared to data collected by the GDHS that states that for children between the ages of 6-23 months, only 47.6% of them consumed meat, fish and poultry 24 hours prior to data collection [37]. As stated earlier, apart from agriculture and forestry, fishing is one economic activity that is common among inhabitants of Upper Manya Krobo district [153]. As a primary economic activity in the communities, fishing can increase accessibility of fish hence increase consumption compared to the Northern region where the inhabitants are predominantly farmers and traders living in mostly landlocked communities [59]. Geographic difference between northern and southern Ghana contributes to the kinds of food being grown, harvested, and hence dietary composition of the foods being consumed [18]. These geographic differences could explain the regional variations of omega-3 index across the two populations. The omega-3 index for Southern Ghanaian children are also higher than values reported for Australian school aged children [160], UK children aged 7-9 years [161], Tanzanian children 2-6 years [64] and in European children 3-8 years [162]. The 96 reasons for these variations is likely attributable to the dietary variation in the various populations. Omega-6 FAs have crucial roles in growth. A previous study from our group reported an association of n-6 PUFAs and linear growth in Tanzanian children [63]. LC n-6 PUFAs such as AA modulate varied physiological responses such as cell growth and differentiation and in conditions that involve altered cellular proliferation [82]. Although, this population had a lower stunting levels, most of the n-6 FAs were significantly lower than those reported in a population whose stunting rates are higher. The significant difference in delta-5 and delta-6 desaturase activity between stunted and non-stunted groups can be attributed to differing levels of DGLA among both groups. As much as a decrease in dietary consumption of n-6 precursor LA in this population may lead to a decrease in n-6 FA metabolites, which could further lead to the suppression of the n-6 FA pathways by the high amounts of dietary n-3 FAs. These findings support previous observations that, the n-3 and n-6 FA pathways are competitive and an increase in dietary intake of one could suppress the other [122]. The exact mechanism of n-3 FAs in growth remains unclear, however their roles in growth is undisputed [163]. Omega-3 FAs have beneficial roles in bone metabolism, by increasing bone formation that affects peak bone mass and reducing bone loss [25]. This is because n-3 FAs reduce inflammatory cytokines, increases calcium resorption and enhances calcium levels [25]; roles that contribute to growth in children. The southern Ghanaian children who participated in this study were reported to have high fish and sea food consumption. Fish and seafood are superior sources of EPA and DHA. They also contain high quality proteins, amino acids, fiber, vitamins and minerals [164, 165]. In 97 addition, fish and seafood have high amounts of essential amino acids such as leucine, lysine and tryptophan [166] which would lead to lower levels of protein energy malnutrition. A recent study by Semba et al. reported that child stunting is associated with circulating amino acids and also reported that stunted children have a limited amount of tryptophan and lysine in their diet. [167]. This study population did not record high levels of stunting. The low level of stunting could possibly be due to high amounts of essential amino acids from high dietary intake of fish and seafood in the population since the children were recruited from fishing communities [153]. The high intake of proteins in this population can rule out protein energy malnutrition which leads to growth stunting. It is also likely that the high intake of proteins in this population may enhance lipid synthesis and further promote linear growth. This is the first study to assess whole blood FAs in Southern Ghanaian children 2-6-year olds. This study utilized whole blood FA biomarkers and the association with growth. The use of a validated dried blood spot collection and blood transport system made the study logistically easier to conduct. This method was also successfully used in a similar study in Northern Ghana [156] and in Tanzania [63]. Aside from these strengths, this study also has limitations that should be noted. We did not collect information on extensive dietary intake hence we are unable to account for the role of other nutrients on growth parameters. Iodine, zinc and other vitamins and minerals have roles to play in growth, but the study could not account for all these nutrients. The collection of blood samples took place throughout the day and the subjects were not required to fast. Since the FA intake across this population is similar, this will not have any effect on the variability of the whole blood FA measurements. Finally, this study cannot be generalized to the whole Ghanaian 98 population because the study was performed in one area in Southern Ghana and dietary intake of foods may differ across locations in Ghana. Conclusions This study assessed whole blood levels of EFAs and their relationship with growth parameters in Southern Ghanaian children 2-6 years of age. EFAD was observed in 10.64% of children while stunting was observed in 22% of all children, yet the was no association observed with whole blood FA and growth parameters. These data suggest that other factors other than EFAD may influence stunting in this population. There is therefore the need for further studies to explore other nutrients and other factors that could be related to stunting in this population. Further, over 90% of all subjects reported dietary intake of fish and seafood prior to blood collection and this could decrease the risk of protein-energy malnutrition and EFAD as a potential cause of stunting in this population. Interestingly, the whole blood n-3 FAs are much higher in this population than that reported in children in other geographical areas. The intake of fish in this population could account for low stunting levels in this population consistent with previous studies demonstrating that increasing consumption of fish and fish oil intake can increase in linear growth in children [165, 167]. There is a need for similar randomized studies, in other populations in Ghana with both high and low fish consumptions. 99 CHAPTER 7: ASSOCIATION OF WHOLE BLOOD FATTY ACIDS WITH COGNITIVE FUNCTION IN 2 TO 6-YEAR-OLD SOUTHERN GHANAIAN CHILDREN. Abstract The role of essential fatty acids (EFAs) and the long chain polyunsaturated FA docosahexaenoic acid (DHA), on cognition and brain development have been demonstrated in a number of studies. The objective of this study was to investigate the relationship between whole-blood FAs and executive function in children from Southern Ghana. A total of 209, 2-to-6-year-old children attempted the dimensional change card sort (DCCS) task to assess executive function, and dried blood spot samples were collected and analyzed for FA content. Of the 209 participants only 25 of them were able to complete the instructional phase. Twenty children were randomly selected again from the same population for re-testing. Among the children who attempted the DCCS test, whole blood levels of linoleic acid was positively associated with executive function. The analysis further showed that, children with low levels of myristic acid (p=0.024) and total saturated fats (p=0.049) performed better in the initial phase of the DCCS test. Children who were picked up from private schools performed better than those in public schools. Children who were picked up at home performed poorly on all phases of the DCCS test. Our study supports that lower levels of whole blood saturated fats may improve cognitive function. Also children who have been exposed to education may perform better at the DCCS test, hence future studies can probe to know the effect of education in DCCS in children in developing countries. 100 Introduction Human growth, both in the fetus and neonates are greatly enhanced by essential fatty acids (EFAs) and their long chain metabolites [85-87]. During pregnancy and early childhood, they accumulate brain and neonatal tissues. [85]. Long chain polyunsaturated fatty acids (LCPUFA) are also concentrated in the central nervous system [88]. They have significant roles in neuronal growth and differentiation of cells and have been associated with cognitive abilities [88-90]. Further, retinal tissues and the brain depend on EFAs, especially for membrane fluidity and signal transduction [26]. LCPUFAs have critical roles in brain development therefore poor PUFA status may affect brain development and cognitive abilities in children [90]. LCPUFA could be included in the diets of infants and children to ensure optimal brain development [92-94]. The diets in most developing countries lack animal foods because majority of inhabitants in such populations are unable to afford diets rich in animal foods [33, 90]. Lack of animal foods may lead to PUFA deficiency[90]. Precisely, the Ghanaian diet are high in carbohydrates but low in fats and proteins[96]. This dietary pattern pre-disposes a section of the population to EFA deficiency. Numerous studies on EFAs have been limited to the use of supplements comprising of EFA [linoleic acid (LA), alpha linolenic acid (ALA)] and/or their metabolites [DHA, eicosapentaenoic acid (EPA) and arachidonic acid (AA)]. The results of these studies have established the roles of EFAs in improved cognition [11, 13, 97, 98]. A study conducted in Tanzania showed that whole blood FA status was associated with cognitive abilities in children 4-6 year old [64]. Similar unpublished studies have been conducted in a population in Northern Ghana, a population with low fish intake. 101 However no study of the association between whole blood FAs and cognitive function has yet been conducted a population with high fish intake. Executive function, the conscious control of thoughts and action involves inhibition, working memory and task switching [100]. Executive function develops in children between the ages of two and ten years [100] and it is controlled by the frontal and temporal lobes of the brain [99]. The frontal and temporal lobes of the brain have high amounts of AA and DHA and they continue to develop after the second year of life [101]. Executive function can be assessed by the dimensional change card sorting (DCCS) task, which is a validated method used to assess cognitive function [100] [102]. This study utilized the DCCS task to assess executive function in Ghanaian children age 2-6. In the Ghanaian population, little is known about cognitive function assessment their association with FAs. We hypothesized that whole blood levels of EPA, DHA, and both EFAs (ALA and LA) would be positively associated with performance on the DCCS test. Methods Study setting The study was conducted in a rural village in the eastern region of Ghana, specifically, in the Upper-Manya Krobo district whose district capital is Asesewa. The entire district has a population of 72,092 and covers 859.1 sq. km. The district comprises of 13,111 households with an average household size of 4.6 persons per household. The temperature ranges from 26°C to 32°C with rainfall ranging from 900mm to 1500mm. The district lies within the semi-deciduous forest and savanna zone. Agriculture, forestry and 102 fishing employs 72% of the workforce aged 15 years and above, constituting the largest industry in the locality. Household water supply is usually from boreholes, tube well, pumps, rivers. Metal sheet is the main roofing material for housing (87.9%). Illiteracy level is high with 33.3% of all inhabitants 11 years and above having no education. The inhabitants in Asesewa, like any community in Ghana are at risk of diseases and other contagious illnesses. The community has one hospital, 3 maternity homes, 4 health centers and 15 Community Health Posts[153] Sample size and subjects Children (n=242) between 2 to 6 years of age residing in communities in the Upper Manya Krobo district, Ghana were recruited for the study. A sub-sample from a larger cohort, recruited as previously described [154] were enrolled in this study. At the community level, the urban town of Asesewa, the district capital was exempted from the study. Further, communities that were inaccessible for more than two weeks during a given period were also excluded. At household level, a household with a target child who had a medical/birth defect that affect eating and normal growth was excluded (eg. cerebral palsy). Data were collected from March to July 2017. Ethical Standards Disclosure All procedures involving human subjects/patients were approved by the Ethics review board at McGill University Canada (IRB#180-1013), the Institutional Review Board at Michigan State University (IRB # 16-557) and the Nogouchi Memorial Institute for Medical 103 Research ethics committee (IRB # 027/13-14). Written informed consent was obtained from all subjects/patients. A script of the written consent was read and translated in Twi, Krobo and Ewe to the parents or caregivers of the children. The parent or caregiver of the participating child gave consent prior to the child’s participation. They were assured that participation was voluntary and confidential, and that their information would remain anonymous. Dietary intake assessment The intake of various foods consumed by subjects within 24 hours of blood sample collection was assessed using a structured questionnaire. The foods that were of interest include fish and seafood, diary, meats and fruits. Anthropometric measurements Anthropometric measurements were taken using averages of the records taken. With a Shorrboard stadiometer (Weigh and Measure LLC, USA), heights of all participants were measured to the nearest 0.1cm. Weight was measured using a digital bathroom scale to the nearest 0.1kg (Tanita BMB-800, Japan). The average of two height and weight measurements were recorded. The date of birth was recorded from the child’s health card or birth certificate. The gender of the child was also recorded. Height, weight, date of birth and sex data were entered into World Health Organization (WHO) Anthro [69] and WHO AnthroPlus [72] software to calculate height-for-age (HAZ), weight-for-age (WAZ), weight-for-height (WHZ), and BMI-for-age (BAZ) z-scores. 104 Blood fatty acid assessment Using a sterile single-use lancet, capillary blood sample (40ul) was obtained by puncturing the middle finger as previously described by Jumbe et al., 2016 [63, 65]. The first drop of blood was wiped with a sterilized dry pad. The drops of blood were then collected onto the DBS cards, cards which are pre-treated with anti-oxidant cocktail. The cards were stored in a dry, cool environment and shipped to the USA for FA analysis at OmegaQuant Analytics, LLC (Sioux Falls, SD). The cards are kept averagely for 8 days, thus from, the sample collection time and time of arrival to the US lab. The samples were stored at –80°C till they were analyzed as previously described [66-68]. Briefly, the cards were punched and combined with the derivatizing reagent [boron trifluoride in methanol (14%), toluene, and methanol (35:30:35 parts)]. The mixture was shaken and heated at 100°C for 45 minutes. Forty parts of both hexane and distilled water were added after the mixture had cooled. The mixture was vortexed and then separated into distinct layers. An aliquot of the hexane layer that contained the FA methyl esters was extracted. FA analysis was performed as previously described [70, 71, 105]. Unless otherwise stated, whole blood FA proportions are expressed as a percent of total identified FAs. Hemoglobin and malaria status Using a hemocue photometer (HemoCue 301, Angelholm, Sweden) hemoglobin concentration was determined. The malaria status was determined using an antigen- based malaria rapid diagnostic test (RDT) kit (Standard diagnostic Inc., Korea). These tests were conducted using additional drops of blood from the same punch. 105 Cognitive assessment: Dimensional change card sort (DCCS) The DCCS [100, 106] is a cognitive function assessment tool that requires the child sort a series of bivalent cards based on one of two instructed dimension (i.e., either color or shape).Initially, the child is asked to sort a series of eight cards based on colour. The child is instructed to switch the categorization dimension and sort another series of eight cards based upon shape (figure 3), after sorting an initial series of eight cards based upon color, This dimensional change in sorting behavior offers an index of executive function as the child must subdue their previously learned set of rules (i.e., sorting by color) and attentional inertia towards those attributes in order to flexibly adjust their behavioral actions and attention to sort the cards by a new set of rules (i.e., sorting by shape) [100, 107]. The child was considered to have passed if he/she correctly sorted 6 of the 8 cards in both the pre- and post-switch phases of the task for each level of the DCCS test. Considering the population of interest and the large developmental spectrum assessed, four levels of the DCCS test were utilized to ensure a vigorous assessment of executive function. While children who passed the first (instructional) level were allowed to take other 3 levels, children who failed (scored less than 6 out of 8) the first level were considered to not be able to follow instructions and not allowed to take other levels of the DCCS test. An initial condition was performed to assess if the child’s executive function was sufficiently developed to enable them to follow directions, this was important because previous research has demonstrated that children younger than 48 months of age particularly struggle to complete this task,[64, 107, 108]. This condition utilized the same pre- and post-switch procedure as outlined above but utilized monovalent cards that only presented a singular dimension (i.e., either color or shape). If the child was able to pass 106 this initial condition, they were then asked to attempt three additional conditions of the DCCS test. These conditions replicated the traditional DCCS test using bivalent cards, but manipulated the attentional characteristics of the cards by progressively integrating the color and shape attributes to reduce practice effects (see Figure 1)[64]. The total number of DCCS test conditions passed was used as an index of executive function[64]. The mother or caregiver was present during all conditions of the DCCS test to observe the process and allow the child to feel comfortable and confident. Of the 242 children who were recruited, only 25 of them were able to complete the instructional test. This was unjustified because most of the children (n=149) were aged 3 years and above at the time of the testing. A number of reasons were speculated for the poor performance of the DCCS test. Among the reasons included inconsistencies on how the DCCS test was administered. In view of this, 20 subjects were randomly selected and the test was administered again. Data reduction and statistical analyses Z-scores for the growth parameters HAZ, WAZ, BAZ and WHZ were calculated using the WHO Anthro [69]. Means and standard deviations were calculated for descriptive analysis. Based on the WHO standard population and definitions of moderate and severe stunting, wasting, and underweight [74], stunting percentages were calculated. The FA values presented here are expressed as percent composition of total blood FAs. Total n- 3 FA proportions were calculated as ∑ [alpha-linolenic + eicosapentaenoic acid (EPA) + docosapentaenoic n-3 + docosahexaenoic acid (DHA)]; total n-6 FA proportions were 107 calculated as ∑ [linoleic + linoelaidic + eicosadienoic + dihomo-gamma-linolenic + arachidonic + docosatetraenoic + docosapentaenoic n-6]; total n-9 FA proportions were calculated as ∑ [oleic + elaidic + eicosenoic + Mead + nervonic]; total saturated FA proportions were calculated as ∑ [myristic + palmitic + stearic + arachidic + behenic + lignoceric]; total MUFA proportions were calculated as ∑ [palmitoleic + oleic + palmitelaidic + nervonic + elaidic + eicosenoic]. T/T ratio was calculated from the ratio of Mead acid and AA[29]. FA product/precursor ratio was used to estimate the desaturase activity [155] as follows: D5D= AA/DGLA; D6D= DGLA/LA. The analyses presented here are for the 25 participants who were able to complete the instructional phase at the first testing and the 20 subjects who were re-tested. Descriptive analyses were conducted to obtain means and standard deviations for all participants. Means between groups (i.e. those who passed the initial condition of the DCCS test versus those who did not pass were compared using t-tests (for continuous data). Models for linear regression included the FA of interest, and covariates hemoglobin, age, malaria and BMI-for-age (BAZ). Hemoglobin concentration was included in our model as a confounder because in similar populations it is a significant predictor of cognitive abilities [110]. Age, malaria and BAZ also showed significant association with the dependent variable (total passes). Pearson and Spearman correlations were performed on the data. SPSS version 24 were used for data analysis. 108 Results Subject characteristics The data collection for this study was done twice due to the reasons stated below. Cognitive function assessment with the DCCS showed that 25 out of the 209 children (12%) were able to pass the initial condition. This means that majority of the children (88%) were unable to proceed to the next phases of the test, based to the conditions of the testing. This was lower than expected. A number of reasons were speculated, and the most probable one was attributed to the way the cognitive function test was administered. The team decided to re-test some of the children to test the new hypothesis: whether the children’s inability to complete the DCCS test was due to the way the test was administered. The second testing showed that, of 18 children who were unable to do the test during the first testing, 7 (38.8%) of them were able to complete the initial phase, while 11 (61.2%) were unable to complete the initial phase, hence couldn’t complete the test. Due to the reasons stated above, the results presented in this section can be grouped into three:  The subjects who passed the initial phase at the first testing, n=24.  Merged data of subjects who passed the initial phase and those who did not pass, n=209  Subjects who passed the initial phase at the second testing, n=20 109  Merged data of subjects who passed the tests at both testing, n=38 (Some of the children, (n=3), participated in both testing and there was missing data for 2 participants. Table 25 presents subject characteristics for all children who passed the initial instruction. The average age of all the children who passed the initial phase is 45.6 months with a range of 38.0-51.0 months. The mean hemoglobin concentration is 11.0 and was within the normal range [111]. The mean weight and height is 14.2 kg and 97.2cm respectively. Z-scores were used to calculate the prevalence of stunting (HAZ), malnutrition (BAZ) and wasting (WAZ). According to WHO standards,[112]. Over 90% of the 24 participants had normal HAZ, WAZ and BAZ scores. 110 Table 25: Characteristics of children who passed the initial instruction (n=24) Mean SD Range Age (mo) Sex (male) n % Malaria (%) Height (cm) Weight (kg) HB (g/dL) BAZ 45.9 10 41.7 4.2 97.2 14.2 11.0 -0.30 10.14 38.0-51.0 4.5 1.6 1.2 0.94 87.9-105.5 10.5-17.7 6.9-12.4 -2.05-1.83 HAZ WAZ HB, hemoglobin; BAZ, BMI-for-age z-score; HAZ, height-for-age z-score; WAZ, weight-for-age z-score. The WHO definitions of moderate and severe stunting, wasting, underweight and malnutrition were applied to the data [168] -2.2-1.87 -2.8-0.85 -1.07 -0.88 0.88 0.82 Children who passed the initial condition of the DCCS test were found to be older, taller and heavier than children who failed the initial DCCS test (Table 26). Of the 209 children who attempted the DCCS task, 24 children (11.5%) were able to follow directions as indicated, and passed the initial condition of the DCCS. Five (5) children (2.4%) passed the initial condition but not any other DCCS conditions, 6 children (2.9%) passed two DCCS conditions, 6 children (2.9%) passed three DCCS conditions, and 7 children (3.3%) passed all four DCCS conditions. When the FA levels between the two groups were compared, children who passed the initial conditions reported significantly lower levels of myristic acid (p=0.024), ALA (p=0.01) nervonic acid (p=0.04) and total saturated FAs (p=0.049). 111 Table 26: Characteristics of children stratified by dimensional change card sort performance for the initial condition (Mean values and standard deviations; numbers and percentages) Pass Fail P <0.001 0.001 <0.001 0.386 0.789 0.505 0.089 0.566 0.024 0.805 0.049 0.246 0.239 0.653 0.668 0.011 0.218 0.770 0.337 0.09 0.04 0.06 0.074 Anthropometry SFAs Omega-6 FAs Omega-3 FAs Others Age (mo) Age range Sex (male) n % Height (cm) Weight (kg) Malaria (%) HB (g/dL) BAZ HAZ WAZ Myristic Palmitic Total SFA1 LA DGLA AA Total n-62 ALA EPA DHA Total n-33 n-6/n-3 Nervonic acid Eicosadienoic (n=24) (n=185) Mean ± SE 45.92±0.67 37.32±0.75 38.0-51.0 13.0-52.0-69.6 10 87 41.6% 47.0% 97.3±0.88 90.69±0.0.52 14.12±0.34 1.96±0.04 12.6±0.16 4.96±1.24 10.96±0.24 10.88±0.10 -0.40±0.20 -1.06±0.17 -0.93±0.17 0.56±0.05 25.57±0.38 39.56±0.24 17.11±0.31 -0.25±0.07 -1.38±0.07 -1.05±0.07 0.85±0.05 25.66±0.11 40.22±0.11 16.63±0.14 1.19±0.0.05 1.24±0.02 9.04±0.36 9.20±0.0.11 29.29±0.56 26.06±0.19 0.20±0.02 0.71±0.07 5.03±0.24 6.90±0.33 4.46±0.21 0.71±0.07 0.23±0.01 0.26±0.01 0.81±0.03 5.10±0.06 7.18±0.09 5.10±0.07 0.84±0.02 0.25±0.01 Oleic 21.89±0.70 20.85±0.19 HB, hemoglobin; BAZ, BMI-for-age z-score; HAZ, height-for-age z-score; WAZ, weight-for-age z-score. The WHO definitions of moderate and severe stunting, wasting, underweight and malnutrition were applied to the data [72]. 1Total SFA includes myristic, palmitic, arachidic, behenic, lignoceric; 2Total n-6 includes linoleic, linolaidic, ϒ-linolenic, eicosadienoic, di-homo-gamma-linolenic, arachidonic, docosatetraenoic, docosapentaenoic n-6; 3Total n-3 includes alpha linolenic, eicosapentaenoic, docosapentaeonic n-3, and docosahexaenoic; 112 Regression between fatty acids and executive function measures Regression analysis between selected FAs and DCCS performance, (n=209) adjusting for age, malaria, BAZ and Hb, is shown in Table 27. Oleic acid (p=0.05) and LA (p=0.03) were positively associated with DCCS performance. To test the hypothesis that whole blood levels of EPA, DHA, and both EFAs (ALA and LA) would be positively associated with executive control as indexed by performance on the DCCS tasks, multiple linear regression using EPA, DHA, ALA, LA, Hb, malaria age, and BAZ was conducted. The model explained 11% of the variation (r2=0.112; adjusted r2=0.077, p0.002). LA (β=0.069, p=0.034) was the only significant contributors to the model. A full model including all 25 single FAs as well as Hb concentrations, malaria, age, and BAZ was significant (p=0.003) and explained about 24% of the variance (r2=0.239; adjusted r2=0.121). In the full model, none of the FAs were associated with DCCS performance. 113 Table 27: Regression results for performance on the dimensional change card sort test and selected fatty acids (FA). (Model: Total pass= Fatty acid of All Children + Age + BAZ + Hemoglobin+malaria) Class Regression results for all subjects (n=209) Fatty acid SFA n-3 FA n-6 FA n-9 FA Desaturases Others Myristic Palimitic Stearic acid Behenic Lignoceric Arachidic Total SFAa Alpha-linolenic Eicosapentanoic Docosahexaenoic DPA Omega-3 Index Total n-3b Linoleic Arachidonic GLA DGLA Docosatetraenoic Total n-6c Mead acid Oleic Eicosenoic Nervonic Elaidic Total n-9d SCD n-7 SCD n-9 D6D D5D Total MUFA Total PUFA T/T ratio n-6/n-3 ratio B ± SE -0.14 ± 0.11 -0.00±0.04 -0.10±0.05 -0.69±0.41 -0.33±0.18 -0.16±0.95 -0.10±0.04 1.03±0.63 -0.09±0.17 -0.04±0.06 -0.63±0.29 -0.03±0.05 -0.05±0.05 0.07±0.03 0.03±0.04 -1.29±0.66 -0.58±0.27 -0.12±0.24 0.02±0.02 -1.25±1.1 0.04±0.02 0.58±0.44 -0.41±0.20 0.47±0.496 0.04±0.02 -10.38±3.86 0.35±0.16 -10.95±3.59 0.04±0.04 0.03±0.02 0.00±0.02 -5.82±9.31 0.17±0.07 P 0.20 0.98 0.06 0.09 0.07 0.86 0.01 0.10 0.61 0.48 0.03 0.47 0.27 0.03 0.45 0.05 0.03 0.61 0.40 0.26 0.05 0.18 0.05 0.34 0.08 <0.01 0.03 <0.01 0.27 0.22 0.84 0.53 0.02 Palmitoleic -0.375±0.14 0.01 Eicosadienoic -2.22±1.17 0.06 Palmitelaidic -1.87±1.26 0.14 aTotal SFA includes myristic, palmitic, arachidic, behenic, lignoceric. bTotal n-3 includes alpha-linolenic, EPA, DPA n-3, and DHA. cTotal n-6 includes linoleic, linolaidic, GLA, eicosadienoic, DGLA, arachidonic, DTA, DPA n-6. 114 Table 27 (cont’d) dTotal n-9 includes oleic, elaidic, eicosanoic, Mead, nervonic. SFA, saturated fatty acid; SCD n-7, stearoyl CoA desaturase n-7; SCD n-9, stearoyl CoA desaturase n-9; D6d, delta-6-desaturase; D5d, delta-5-desaturase. Regression analysis between selected FAs and DCCS performance, (n=24) adjusting for age, malaria, BAZ, Hb is shown in table 28. Total MUFAs and total n-9 were positively associated with DCCS and they were trending towards statistical significance. 115 Table 28: Regression results for performance on the dimensional change card sort test and selected fatty acids (FA). (Model: Total pass= Fatty acid of All Children + Age + BAZ + Hemoglobin+malaria) Class Regression results for total pass (n=24) Fatty acid B ± SE P SFA n-3 FA n-6 FA n-9 FA Desaturases Others Myristic Palimitic Palmitelaidic Stearic acid Elaidic Behenic Arachidic Total SFAa ALA EPA DHA DPA Omega-3 Index Total n-3b LA AA GLA DGLA DTA Eicosadienoic Total n-6c Mead acid Oleic Eicosenoic Nervonic Total n-9d SCD n-7 SCD n-9 D6D D5D Total MUFA Total PUFA T/T ratio -0.76 ± 0.88 0.14±0.13 -2.35±4.67 -0.10±0.21 1.34±1.74 -0.91±2.28 2.00±4.65 -0.18±0.21 -0.45±3.50 -0.26±0.73 -0.23±0.18 -1.27±0.93 -0.16±0.14 -0.17±14 -0.14±0.17 -0.23.±0.12 -6.51±3.34 -0.96±0.88 -0.18±1.23 -4.79±4.33 -0.15±0.08 -4.54±5.56 0.11±0.06 -0.14±2.65 -0.21±1.17 0.12±0.07 -10.09±33.40 0.55±0.43 -9.67±13.86 -0.12±0.19 0.12±0.06 -0.09±0.05 2.79±46.64 0.40 0.31 0.62 0.61 0.45 0.69 0.67 0.41 0.90 0.73 0.23 0.18 0.27 0.24 0.41 0.08 0.06 0.29 0.89 0.28 0.08 0.42 0.10 0.96 0.86 0.08 0.77 0.22 0.49 0.53 0.08 0.11 0.95 n-6/n-3 ratio aTotal SFA includes myristic, palmitic, arachidic, behenic, lignoceric. bTotal n-3 includes alpha-linolenic, EPA, DPA n- 3, and DHA. cTotal n-6 includes linoleic, linolaidic, GLA, eicosadienoic, DGLA, arachidonic, DTA, DPA n-6. dTotal n-9 includes oleic, elaidic, eicosanoic, Mead, nervonic. SFA, saturated fatty acid; SCD n-7, stearoyl CoA desaturase n-7; SCD n-9, stearoyl CoA desaturase n-9; D6d, delta-6-desaturase; D5d, delta-5-desaturase. 0.20±0.22 0.39 116 Regression analysis between selected FAs and DCCS performance, (n=38) adjusting for age, malaria, BAZ, Hb is shown in Table 29. None of the FAs were positively associated with DCCS test. 117 Fatty acid B ± SE Regression results for total pass(both testing) (n=38) -0.93 ± 0.85 0.06±0.13 -1.87±3.75 -0.15±0.16 1.31±2.07 -0.14±0.17 2.88±2.63 0.56±0.65 -0.03±0.18 -0.94±0.92 0.02±0.15 0.00±14 0.05±0.16 -0.12±0.13 -3.17±3.24 -1.41±0.73 -0.85±0.97 -6.69±4.63 -0.06±0.09 -3.24±4.03 0.06±0.06 -2.16±1.34 0.06±0.07 Table 29: Regression results for performance on the dimensional change card sort test and selected fatty acids Class SFA n-3 FA n-6 FA n-9 FA Desaturases Others Model: Total pass= Fatty acid of All Children + Age + BAZ + Hemoglobin+malaria aTotal SFA includes myristic, palmitic, arachidic, behenic, lignoceric. bTotal n-3 includes alpha-linolenic, EPA, DPA n-3, and DHA. cTotal n-6 includes linoleic, linolaidic, GLA, eicosadienoic, DGLA, arachidonic, DTA, DPA n-6. dTotal n-9 includes oleic, elaidic, eicosanoic, Mead, nervonic. SFA, saturated fatty acid; SCD n-7, stearoyl CoA desaturase n-7; SCD n-9, stearoyl CoA desaturase n-9; D6d, delta-6-desaturase; D5d, delta-5-desaturase. Myristic Palimitic Palmitelaidic Stearic acid Elaidic Total SFAa ALA EPA DHA DPA Omega-3 Index Total n-3b Linoleic Arachidonic GLA DGLA Docosatetraenoic Eicosadienoic Total n-6c Mead acid Oleic Eicosenoic Total n-9d SCD n-7 SCD n-9 D6D D5D Total MUFA Total PUFA T/T ratio n-6/n-3 ratio 0.12±0.19 0.06±0.06 -0.03±0.06 -6.31±28.81 0.03±0.24 P 0.28 0.67 0.62 0.36 0.53 0.41 0.28 0.40 0.91 0.31 0.91 1.00 0.75 0.37 0.34 0.06 0.39 0.16 0.47 0.42 0.32 0.13 0.35 0.62 0.30 0.06 0.32 0.38 0.62 0.83 0.89 -10.21±20.13 0.46±0.43 -22.52±11.52 Education and performance in the DCCS test In the second testing some of the children were picked from private schools, public schools and others were picked from home. The information on the educational 118 component is presented in table 30. Of those who were picked from public schools 100% passed the initial condition while 66.7 % passed 2 or more post-switch tests. Of all those were picked from public schools, 80% of them passed the initial test while 60% passed 2 or more post-switch test. Of all those who were picked at home only 25% of them were able to pass the initial test, while none of them were able to pass 3 or more of the post tests. Table 30: Educational component Passing the initial test Passes 2 or more post-tests Passing 3 or more post tests Children picked from private schools (n=3) Children picked from public Children picked at schools, n=5 home, n=12 100% 66.70% 66.70% 80% 60% 40% 25% 33.30% 0 Table S4 to S6 presents Spearman correlation and Pearson correlation coefficient and p values for the subjects, n=24, and n=38. None of the FAs were associated with a total passes except for DGLA that showed trending significance for Pearson correlation, n=24. Discussion The objective of this study was to characterize the whole blood FA levels of Southern Ghanaian children and determine FA associations with cognition. Based on the Mead acid levels and the T/T ratio, EFAD was found to be low in this population, 10.5%. From the first testing for the cognitive function assessment, 25 out of the 209 children (12%) were able to pass the initial condition. This indicates that majority of the children (88%) were unable to proceed to the next phases of the test based to the conditions of the test. This was lower than expected. A number of reasons were speculated, and the most probable one was attributed to inconsistencies in administering the cognitive function test. 119 A second testing was scheduled for 20 subjects and this time, with a supervisor who could give further instructions when the test was being administered. The second testing showed that, out of the 18 children who were unable to do the test during the first testing, 7 (38.8%) of them were able to complete the initial phase, while 11 (61.2%) were unable to complete the initial phase. These results show that, the way the test was administered could result in the low pass rate in the first testing. A lack of proper supervision of the interview team could have caused this. The initial phase of the DCCS test requires much sensitivity to get a child settled for the entire procedure. Due to environmental factors, the children may be jittery and not pay much attention to the testing, especially during the initial phase. This may require a tactful investigator to get the accurate scores for the child. An investigator’s inability to maintain calmness in a child can lead to poor performance. Further, when an investigator is burdened during questionnaire administration it may prevent him/her from actively calming down a child to perform the test. The study sample was a sub-sample from a bigger study[154], and the investigators were tasked to take lots of information from the parents of the subject. Additional information could have increased investigator burden, hence the results obtained. This can be prevented in other studies in a number of ways: 1) Supervising the investigators to patiently assess the child can be helpful. 2) Also if possible merging a number of studies can be prevented so that the investigators are not burdened. The data were analyzed regardless to these pitfalls. Our data support the hypothesis that children with higher whole blood levels of EFAs (LA and ALA), as well as DHA and EPA, are more likely to pass the DCCS test, however, in individual regression analysis, LA was 120 positively associated with DCCS performance, which is consistent with results from a previous study in Tanzania [64]. The results also showed that, the children who passed the initial condition of the DCCS test were found to be older and taller, and this is consistent with data obtained from a similar study in Tanzanian children[64]. High levels of saturated fats have been linked with poor cognition because saturated FAs can lead to saturated fats-induced oxidative stress and reduced brain-derived neurotrophic factor (BDNF), resulting in compromised synaptic plasticity. Further, saturated fats increases insulin resistance in the brain and diminish the integrity of the blood brain barrier [169]. These indicate that lower levels of saturated fats may improve cognition and our study showed that children with lower levels of total saturated fatty acids and myristic acid passed the initial phase of the DCCS testing, consistent with studies by Khan et al., 2015[169]. The role of education/schooling of DCCS were assessed. In the second testing, of the 3 children who were chosen from private schools, all of them passed the initial condition while those chosen from public schools had fewer of them passing the DCCS. Children who have not had any education, thus, those chosen from home, performed poorly in all phases of the cognitive function test. This could indicate that being in school can have an effect on attention, task switching and overall executive function of the children. From the onset, this study was presented with a number of pitfalls, among them include the irregularities in the administering of the test. This pitfall contributed to a lower sample size for this study. However, the second testing provided a new perspective to the entire study: the educational component. Although the study was not well powered, the results 121 of this study add to a body of knowledge that high whole blood saturated fats are associated with lower cognitive performance. Future studies should include supervisors when administering the DCCS test so that investigators would be guided to do the right thing. Further, considering the educational level or whether the child has been exposed to any education at all can be assessed in future studies. Finally, measures should be taken to reduce investigator burden so as to get very accurate results and scores. 122 CHAPTER 8: QUANTIFICATION OF FATTY ACID AND MINERAL LEVELS OF SELECTED SEEDS, NUTS, AND OILS IN GHANA. This chapter has been published in: Adjepong et al., Quantification of fatty acid and mineral levels of selected seeds, nuts, and oils in Ghana. JFAC. 2016, Feb 59 (2017) 43–49 Abstract Background: Fatty acids (FA) and minerals play crucial roles in growth and development. However, Ghanaian diets consist mainly of starchy roots and cereals, with intake of fats and many minerals below recommended levels. The purpose of this study was to quantify FA and mineral levels of seeds, nuts, and oils in Ghana that are available but not usually incorporated in the diets. Method: Seven seeds and five oils collected in Ghana were analyzed for FA and mineral composition by gas chromatography mass spectrometry (GC/MS) and inductively coupled plasma (ICP) emission spectroscopy, respectively. Results: Soybean was found to contain high levels of alpha-linolenic acid (ALA) (3.77 mg/g). Linoleic acid (LA) was higher in peanuts (65.8 mg/g), agushie seeds (102 mg/g) and agushie flour (122 mg/g). Agushie seeds (88.4 mg/g), agushie flour (111 mg/g) and soybean (78.4 mg/g) had appreciable levels of iron. In addition, both agushie seeds and flour contained high amounts of zinc. Taken together, these data indicate that several Ghanaian seeds, nuts, and oils are high in FA, including essential fatty acids, and minerals. Future studies should investigate increased incorporation of palm oil, soybean, peanuts, cashew nut, tigernuts, agushie seeds and/or flour into Ghanaian diets in areas where nutrient deficiencies are prevalent. 123 Introduction Essential fatty acids (EFA) and minerals play crucial roles in human growth and development due to their role in nutrient metabolism and cell differentiation [8, 170]. FA are precursors of structural and metabolic substances in the body as well as a major contributor of energy in diets. Although fats have critical roles in human diets, their intakes in African diets are low [33]. Ghanaian diets consist mainly of starchy roots (yam and cassava), fruits (plantain, orange) and cereals (rice, maize), which together make up about 75% of diets, while protein and fats make up 25% [134]. This indicates how it is generally accepted that there is low dietary diversity in Ghana often below adequate levels of dietary fats and proteins. In addition, only 36.6% of Ghanaian children from 6 to 24 months old have fats added to their complementary food [171]. The low content of fats in Ghanaian diets contributes to lower intakes of EFA, a probable likelihood of EFA deficiency (EFAD) in the population. Mineral deficiencies have also been reported in Ghana [134], contributing to the occurrence of iron deficiency anemia. EFAD is related to stunting and cognitive impairment, hence there is a need to include adequate amounts of EFA and minerals in diets. The inclusion of EFA in the diet is of utmost importance because EFA cannot be synthesized in the body and must be provided through the diet. They are abundant in foods such as soybean oil, canola oil, flaxseed oil, nuts, and animal products. They include linoleic acid (LA), an n-6 FA, and alpha-linolenic acid (ALA), an n-3 FA. Downstream products of EFA, such as arachidonic acid (AA), are metabolized from LA, while eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are metabolized from ALA. EPA and DHA are predominantly found in seafood. The EFA and their very 124 long chain polyunsaturated fatty acid (VLCPUFA) derivatives are vital for growth, cognition, and immune functions [33, 130]. Fish are a good source but they are expensive, and they are not available in sufficient amounts in the population. This is evidenced by the fact that the amount of fish required to meet the nutritional needs of Ghanaians exceeds fish production (marine, inland and aquaculture) [131]. There is a high probability of low intakes of EFA in the Ghanaian population; therefore, the exploration of alternative and affordable sources of EFA is important. The prevalence of mineral deficiencies in different age groups has been described [34, 129, 133]. The prevalence of anemia in Ghanaian children 6 to 59 months old is 78% and iron deficiency is a major contributor [19]. A study of a community in Ghana showed that zinc deficiency among children 2 to 10 years was 35.6% [133]. Mineral deficiencies are also known to be associated with growth stunting and cognitive impairment [135, 136]. This is because some minerals such as iron and zinc are cofactor nutrients and are beneficial in EFA metabolism. For example, zinc deficiency can impair the conversion of LA to AA and ALA to EPA and DHA [137]. Iron deficiency also contributes to the occurrence of iron deficiency anemia. Studies have been conducted to evaluate the mineral composition of seeds and nuts, but these data can vary due to environmental factors such as soil type, agronomic practices and climate [138, 139]. Differences between geographical regions can directly affect the concentration of minerals in crop plants and affect dietary mineral content, impairing EFA metabolism and consequently leading to growth stunting and cognitive impairment in the population. However, the mineral content of foods in Ghana has been poorly described. 125 The West African food composition table published in 2012 gives an estimation of nutrients such as proteins, carbohydrates, total fat, vitamins, and minerals, with no information on the EFA concentrations of individual foods and no information on the provenance of the foods analyzed. These reports state that the data represent average values derived from compositional data from 8 countries including Ghana. Additionally, it has been documented that most of the mineral and vitamin data in the table were based on information obtained from several non-African countries. Cultivation of crops in different geographical areas is also known to affect composition of food [149]. Given these drawbacks, the nutrient content listed in the table does not represent the foods in Ghana. Therefore, analyses that quantifies individual FA, especially EFA, and mineral content of local foods in Ghana is needed. In this study, we quantified the amount of EFA, VLCPUFAs and minerals in local foods available in markets in Southern Ghana. Materials and Methods Preparation of selected seeds, oils, and nuts in Ghana. Seeds, nuts, and oils (Table 31) were purchased from the Kumasi Central Market in Kumasi, Southern Ghana. To prevent the oils from oxidation they were packed in plastic containers. The seeds and nuts were crushed and freeze-dried and also stored in containers. The freeze-dried samples were stored no longer than 14 days upon arrival in the US. All samples were shipped to a Michigan State University Food Science and Human Nutrition laboratory. The samples were purged with high purity nitrogen and stored at -20°C until analysis. Freeze-drying of all solid products was chosen to maintain integrity of the foods during shipment due to shipping and power challenges in this region. 126 Table 31: List of seeds, nuts, and oils analyzed Food Soybean Cashew nut Agushie flour Peanut Tigernut Coconut Agushie seeds Palm kernel oil Margarine Vegetable oil Coconut oil Palm oil Sheabutter Abbreviation SYBN CNUT AGSF PNUT TNUT CCN AGS PKO MG VGT CCNO PALM SBUT Genera Glycine sp. Anarcardium sp. Cucumeropsis sp. Arachis sp. Cyperus sp. Cocos sp. Cucumeropsis sp. Elaeis sp. NA Elaeis sp. Cocos sp. Elaeis sp. Vitellaria sp. Description A legume, bean seeds, freeze-dried A tropical tree with nuts. Nuts were freeze-dried Climbing vine, flattened seeds. Seeds milled. A root of a tropical legume. Seeds were freeze-dried A root nut of a tropical ‘grass’. Nuts were freeze-dried A palm tree fruit, flakes were freeze dried and used in analyses Climbing vine, flattened seeds. Seeds were freeze dried Oil derived from the kernel of the oil palm A commercial butter substitute made from hydrogenated vegetable oils Commercially refined oil made from crude palm oil Crude oil extracted from the flesh of mature coconut Crude oil processed locally, red-orange, high in beta carotene An ivory-coloured fat extracted from African shea nut. 127 These freeze-drying methods have been shown to give comparable extraction yields when compared to air-drying methods [141]. Also, LA and ALA from freeze-dried samples are reported to be more abundant than the air-drying methods [142]. Crude Seed Oil Extraction Glassware for the analysis were acid washed to deactivate glassware and remove mineral residues, followed by high performance liquid chromatography (HPLC)-grade organic solvents to remove any FA contaminants. Lipids were extracted from seed material as previously described [143], but modified as specified [172]. In brief, a total of 400 mg freeze-dried seed material was incubated at RT in 10 mL of 2:1 (v/v) mixture of HPLC grade chloroform (Avantor Performance Materials, Inc., Center Valley, PA) and HPLC-grade methanol containing 100 µg butylated hydroxytoluene (BHT)/mL (Sigma – Aldrich, St. Louis, MO). Using lipid-free filters (FGE Healthcare UK Limited, Buckinghamshire, UK), the seed-solvent mixtures were gravity-filtered into glass tubes, containing 2.5 mL of 0.88% m/v aqueous KCl to remove water soluble pigments and phytochemicals (J.T. Baker, Phillipsburg, NJ). After drying under high-purity nitrogen, the total crude seed oil was weighed and calculated. Methylation of Oils to FAMEs, Neutralization, and FAME Isolation A total of 80 mg of the crude seed oil were weighed into individual 16x100 mm glass tubes. The samples were resuspended in chloroform/methanol (2:1 v/v, 100 µg BHT/mL) to obtain a final total lipid concentration (20 mg/mL). The resuspended oils were prepared for methylation as described by Cequier-Sanchez et al. [143]. In brief, 100 µL lipid extract 128 solution was transferred to a clean 16×100 mm Teflon-lined screw-capped glass tubes. To each sample, an internal standard, nonadecanoic acid (150 µg; Sigma – Aldrich, St. Louis, MO) in HPLC-chloroform was added. The resultant mixture was dried under high- purity nitrogen at RT. The samples were methylated with 2% acidified methanol as described methods by Agren et.al. [145], and as modified by Pickens et al. [70]. FAMEs were neutralized and isolated as previously described [71]. FAME Identification, Analysis, and Data Processing Re-suspended FAMEs were transferred to GC vials with glass inserts for analysis. Prior to analysis, the sample injection order was randomized. FAMEs were identified and quantified using a dual stage quadrupole (DSQ)II quadrupole GC/MS (Thermo Scientific, Waltham, MA) equipped with a DB-23, 30-m column (0.25 mm id; Agilent Technologies, Santa Clara, CA) using helium as a carrier gas. The GC temperature profile was as follows: Initial, hold 40°C 1 min; Ramp 1, 100°C/min to 160°C; Ramp 2, 2.8°C/min to 192°C; Ramp 3, 0.5°C/min to 201°C; Ramp 4, 50°C/min to 250°C and hold for 1 min. Selective ion monitoring (SIM) was employed for enhanced sensitivity. Identification and quantification of individual FAMEs was done with standard FAME mixture (Part# CRM47885; Lot# LC06601V; Supelco, Bellefonte, PA). The certified range of each FAME measured was as follows: 600 µg/mL for palmitic acid; 400 µg/mL for myristic, stearic, (cis-9) oleic, arachidic, behenic, and lignoceric acid; 200 µg/mL for (cis-7) palmitoleic, (cis-11) eicosenoic, (cis-9,12) LA, (cis-6,9,12) gamma-linolenic acid (GLA), (cis-11,14) eicosadienoic, (cis-5,8,11,14) AA, (cis-13,16) docosadienoic, (cis-15) nervonic, (cis- 11,14,17) ALA, (cis-5,8,11,14,17) EPA, and (cis-4,7,10,13,16,19) DHA. Standard curves 129 were created from the certified ranges given using 5-fold serial dilutions to produce a 5- point standard curve. Detected FAME concentrations below the lower limit of quantification (LLOQ) are defined for each FA in Tables 2-5. DHA, EPA, and linoelaidic acid were below the LLOQ in all samples analyzed and were excluded from the tables. Resuspended FAME samples that contained FAME concentrations above the highest standard curve were diluted 1:10 and reanalyzed on the same standard curve as undiluted samples. FAME peak integration and quantification was performed using TargetLynx V4.1 (Waters, Milford, MA) based on the FAME standard’s retention time and SIM ions and ratios. The concentrations of resuspended FAMEs were normalized to the amount of food-grade oil (i.e. palm oil) or crude seed oil and total seed material for the seed samples (i.e. agushie seeds). FA concentrations are representative of a pooled food-grade oil (n=1) or pooled seed (n=1) sample, obtained from a local market in Kumasi, Southern Ghana. Figures were made in GraphPad Prism version 4 (GraphPad Software, Inc., La Jolla, CA). Mineral analysis For the mineral analysis, freeze-dried samples of PNUT, AGS, AGSF, CNUT, CCN, TNUT and SYBN were shipped on dry ice to a third-party contractor. The minerals that were analyzed include zinc, iron, potassium, phosphorus, sodium, magnesium, manganese and calcium. Identification and quantification of minerals was performed using ICP emission spectroscopy (ICP_S: 28) AOAC International no. 984.27, 985.01, and 2011.14. [173, 174] For each sample, concentrations of minerals are expressed as mg/kg. 130 Results The composition of FA in the seeds, nuts, and food-grade oils are reported as the concentration of each FA as mg/g (Table 33–36). Samples were found to contain varying levels of several FA, as shown in the following data. Saturated Fats All samples analyzed were found to contain saturated FA. CCNO, PALM and VGT had high amounts of total saturated FA (112 mg/g, 184 mg/g, and 165 mg/g, respectively). In particular, PKO and CCNO had high amounts of myristic acid, while PALM and VGT contained high amounts of palmitic acid. SBUT contained high amounts of stearic acid (Table 32). Monounsaturated Fatty Acids Food-grade oils were more abundant in saturated fats and MUFA compared to the seeds. The MUFA that were quantified include oleic (18:1), eicosenoic (20:1), and docosenoic (22:1). PALM (202 mg/g) and SBUT (309 mg/g) had high amounts of oleic acid. Levels of eicosenoic and docosenoic acid were low in all samples (Table 33). 131 Table 32: Saturated fatty acids (mg FA/g crude oil or mg FA/g seed)1 Sample ID Soybean Cashew nut Agushie flour Peanut Tigernut Coconut Agushie seeds Palm kernel oil Margarine Vegetable oil Coconut oil Palm oil Sheabutter Myristic C14:0 0.06 0.10 0.14 0.06 0.06 21.8