VITAMIN D INTAKE AND VITAMIN D STATUS IN 5 - 6 YEAR OLD CHILDREN IN VANCOUVER by Betina Feldfoss Rasmussen B.Sc., The Metropolitan University College, Copenhagen Denmark, 2011 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORIAL STUDIES (Human Nutrition) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)

October 2013 © Betina Feldfoss Rasmussen, 2013

Abstract Vitamin D is important in maintaining bone health and has recently been proposed to have additional roles in the immune system and brain development. The estimated average requirement (EAR) and recommended dietary allowance (RDA) for vitamin D established by the Institute of Medicine (IOM) in 2011 is 10µg/day and 15µg/day, respectively. When this study was initiated, little information was available on whether vitamin D intakes below the recommendations in young children result in biochemical evidence of vitamin insufficiency or deficiency. Therefore the aims in this thesis were; to estimate vitamin D intakes in children, and the contribution of natural and fortified foods, and supplements; to determine the proportion of children consuming vitamin D below and above the intake recommendations; to use biochemical measures of plasma 25(OH)D to determine the proportion of vitamin D sufficient, insufficient and deficient children; and to estimate the importance of vitamin D intake and season to the children’s plasma 25(OH)D. This was a cross-sectional design with 200 children from Vancouver BC, aged 5.75 years. Vitamin D intakes were estimated using a food frequency questionnaire and 24 hr dietary recalls. Plasma 25 (OH)D was determined by HPLC-tandem mass spectrometry. The median vitamin D intake from foods was below the EAR and RDA. The children obtained 85.9% of their dietary vitamin D from supplements and fortified foods and 14.1% from natural food sources. Total median vitamin D intakes in children given or not given supplements was 13.0 (9.0) µg/day and 4.8 (3.7) µg/day, respectively, P< 0.001. Using the FFQ, 51% and 76% of the children did not meet the EAR and RDA for vitamin D, respectively. However, only 4.7% and 19.0% had a plasma 25 (OH)D below 40 nmol/L or 50 nmol/L, respectively. Unexpectedly, only 12.5% of the children who did not meet the EAR during winter months had a plasma 25 (OH)D below 40 nmol/L. The results in this thesis suggest that children depend on supplements and fortified foods to achieve the current vitamin D intake

ii

recommendations. However, despite apparent low vitamin D intakes, few children show biochemical evidence of vitamin D insufficiency, even during winter months.

iii

Preface This thesis presents work conducted by myself, Betina Rasmussen, under the supervision of Dr. Sheila Innis, with ongoing guidance from Dr. David Kitts and Dr. Tim Green at the University of British Columbia. This project was part of a larger research study, examining essential fatty acid requirements to support optimal childhood development and health. The study protocol was designed by principal investigator Dr. Sheila Innis and funded primarily by her funding from the Canadian Institutes of Health Research and Freedom to Discover awards. The project involves team members with complementary expertise, including Kelly Richardson, with graduate training in child developmental assessments, other graduate students, including Kelly Mulder and Brain Wu, and expert laboratory technicians including Roger A Dyer and Janette D King. My roles in this project included conducting parent interviews, including dietary intake data collection using 24 hr recalls and Food Frequency Questionnaires, and collection of dietary data using 24 hour recalls by telephone. I was responsible with Kelly Mulder for entry of all of the dietary data into the nutrient bases, and I estimated the vitamin D intakes for all diet records. I assisted in sample preparation for the analysis of plasma 25(OH)D analysis by LC-MS/MS, which was done by Roger Dyer. I was responsible for summarizing all of the data relevant to this thesis, conducted all of the statistical analysis in this thesis and wrote the thesis with supervision from Dr. Sheila Innis. Sections of this thesis will be submitted for publication as a manuscript to academic journals. The University of British Columbia Children and Women’s Clinical Research Ethics Board approved this study protocol (identifiers: H09-02921 and H09-01633). I was supported from my graduate training by Statens Uddannelsesstøtte (SU) from the Danish government and from grant funds held by Dr Innis.

iv

Table of contents Abstract ......................................................................................................................................... ii Preface ...........................................................................................................................................iv Table of contents ............................................................................................................................ v List of tables ............................................................................................................................... viii List of figures.................................................................................................................................. x List of abbreviations .....................................................................................................................xi Acknowledgements ..................................................................................................................... xii Chapter 1: Literature review .................................................................................................. 1 1.1

Introduction .............................................................................................................................1

1.2

Vitamin D sources ...................................................................................................................2

1.2.1 1.3

Diet, supplements, fortification and cutaneous synthesis ...................................................2 Vitamin D metabolism ............................................................................................................7

1.3.1

Absorption and transport .....................................................................................................7

1.3.2

Hydroxylation of vitamin D in the liver and kidney ...........................................................7

1.3.3

Physiological roles of vitamin D .........................................................................................8

1.3.3.1

Calcium and phosphate homeostasis and bone health ................................................9

1.3.3.2

Emerging roles of vitamin D ....................................................................................11

1.4

Consequences of low vitamin D status ..................................................................................12

1.5

Factors affecting vitamin D status .........................................................................................12

1.5.1

Individual factors affecting vitamin D status ....................................................................12

1.5.2

Environmental and lifestyle factors affecting vitamin D status ........................................13

1.6 1.6.1

Assessment of vitamin D status .............................................................................................14 Vitamin D sufficiency, insufficiency and deficiency ........................................................14 v

1.7

Vitamin D intake recommendations ......................................................................................16

1.8

Current knowledge of vitamin D intake and status in children .............................................17

1.8.1

Canadian community health measures survey cycle 2.2 - 2004 .......................................17

1.8.2

Canadian health measures survey 2007 - 2009 and 2009 - 2011 ......................................18

1.8.3

USA – National health and nutrition examination survey ................................................19

1.8.4

Research reports ................................................................................................................20

1.8.4.1

Canada ......................................................................................................................20

1.8.4.2

U.S.A ........................................................................................................................22

1.8.5

Summary ...........................................................................................................................24

Chapter 2: Study .................................................................................................................... 30 2.1

Purpose ..................................................................................................................................30

2.2

Objectives ..............................................................................................................................31

2.3

Methods .................................................................................................................................31

2.3.1

Design and setting .............................................................................................................31

2.3.2

Inclusion and exclusion criteria and recruitment ..............................................................32

2.3.3

Demographic characteristics .............................................................................................32

2.3.4

Dietary assessments and collection of information on supplement use ............................33

2.3.5

Analysis of dietary intakes ................................................................................................34

2.3.6

Anthropometrics................................................................................................................35

2.3.7

Blood collection, preparation and analysis of 25(OH)D ...................................................35

2.3.7.1 2.3.8

LC-MS/MS ...............................................................................................................35

Statistical analysis .............................................................................................................37

Chapter 3: Results .................................................................................................................. 39 3.1

Subject characteristics ...........................................................................................................40

3.2

Vitamin D intake ...................................................................................................................44

3.2.1

Vitamin D intakes from foods including natural sources and fortified foods ...................48

3.2.2

Vitamin D intakes from supplements................................................................................49 vi

3.2.3

Vitamin D intake from foods and supplements.................................................................50

3.2.4

Vitamin D intakes compared to the EAR and RDA for children ......................................52

3.3

Plasma 25 (OH)D ..................................................................................................................58

3.3.1

Plasma 25 (OH)D and vitamin D intake of children during different seasons..................59

3.3.2

Plasma 25 (OH)D of children given and not given supplemental vitamin D during

different seasons .............................................................................................................................61

Chapter 4: Discussion ............................................................................................................ 63 4.1

Vitamin D intake ...................................................................................................................63

4.1.1

Vitamin D sources .............................................................................................................64

4.1.2

Vitamin D intake from supplements .................................................................................65

4.1.2.1

Total vitamin D intake ..............................................................................................67

4.1.3

Vitamin D intake compared to the EAR and RDA ...........................................................68

4.1.4

Vitamin D intakes assessed using FFQ and 24 hr recalls .................................................69

4.2

Plasma 25 (OH)D ..................................................................................................................71

4.2.1

Season ...............................................................................................................................72

4.2.2

Relationship of vitamin D intake to vitamin D status .......................................................73

4.3

Strengths and limitations .......................................................................................................76

4.4

Future directions ....................................................................................................................78

Chapter 5: Conclusion ........................................................................................................... 79 References ..................................................................................................................................... 80 Appendices ................................................................................................................................... 87 Appendix A Informed consent form...................................................................................................87 Appendix B Socio-demographic questionnaire ..................................................................................95 Appendix C Food frequency questionnaire ........................................................................................97

vii

List of tables Table 1.1

Vitamin D content in foods ..............................................................................4

Table 1.2

Biochemical indicators of vitamin D sufficiency, insufficiency and deficiency ........................................................................................................16

Table 1.3

Summary of the proportions of children given supplements in different studies..............................................................................................................25

Table 1.4

Summary of vitamin D intakes in children in Canada ....................................26

Table 1.5

Summary of vitamin D intake in children in USA..........................................27

Table 1.6

Summary of vitamin D status of children in Canada ......................................28

Table 1.7

Summary of vitamin D status of children in the U.S ......................................29

Table 3.1

Subject characteristics .....................................................................................42

Table 3.2

Child anthropometric measures ......................................................................43

Table 3.3

Energy intakes (kcal/day) estimated using FFQ and 24 hour dietary recalls ..44

Table 3.4

Daily dietary vitamin D intake and vitamin D intake as µg/1000 kcal from foods estimated using the FFQ, single 24 hr recall and three 24 hr recalls. ...47

Table 3.5

Dietary vitamin D intake (µg/day) and proportion of total intake obtained from different foods ........................................................................................48 viii

Table 3.7

Vitamin D intake as µg/day from foods in children given or not given vitamin D containing supplements ..................................................................51

Table 3.8

Total vitamin D intake from diet and supplements or diet alone estimated using the FFQ and three 24 hr recalls for children given or not given supplements. ....................................................................................................52

Table 3.9

Proportion of total daily vitamin D obtained from supplements, fortified food and natural food sources assessed by FFQ .............................................53

Table 3.10

Effect of supplement use and sex on plasma 25 (OH)D in children. .............58

Table 3.11

Vitamin D intake from diet and supplements during different seasons .........59

Table 3.12

Percentage of children classified as vitamin D sufficient, insufficient and deficient, year round and during different seasons .........................................60

Table 3.13

Plasma 25 (OH)D as nmol/L for supplement users and non users divided into seasons .....................................................................................................62

ix

List of figures Figure 1.1

Production of vitamin D3 and D2 ....................................................................... 6

Figure 1.2

Schematic of vitamin D metabolism ................................................................. 9

Figure 3.1

Study flow chart .............................................................................................. 39

Figure 3.2

Inter-individual correlation between energy intake estimated using the FFQ and a single 24 hr recalls, and between energy intake estimated using the FFQ and three 24 hr recalls ............................................................................. 45

Figure 3.3

Amount of vitamin D consumed in natural food sources, fortified foods and supplements with comparison to the EAR and RDA ............................... 54

Figure 3.4

Proportion of children with intakes of vitamin D from diet or diet plus supplements with vitamin D intakes below the EAR or RDA for vitamin D......................................................................................................... 55

Figure 3.5

Proportion of children given or not given vitamin D supplements with a total intake of vitamin D below or above the EAR or RDA ........................... 57

Figure 4.1

Children with vitamin D intakes below or above EAR (10 µg/day) or RDA (15 µg/day) with a plasma 25 (OH)D above or below 40 nmol/L or 50 nmol/L in winter months ................................................................................. 75

x

List of abbreviations 1,25(OH)D 25 (OH)D AI ANOVA BMC BMD BMI CCHS 2.2 CHMS CI CNF CPS CYP P450 CYP2R1 CYP27B1 D2 D3 DBP DIN DRI EAR FFQ FGF23 IOM IU HPLC-MS/MS PTH RDA SD SES SPE SPF UL USDA UV

1,25-dihydroxyvitamin D 25-hydroxyvitamin D Adequate intake Analysis of variance Bone mineral content Bone mineral density Body mass index Canadian community health survey cycle 2.2 Canadian health measures survey Confidence interval Canadian nutrient file Canadian paediatric society Cytochrome P450 Cytochrome P450 2R1 Cytochrome P450 27B1 Ergocalciferol Cholecalciferol Vitamin D binding protein Drug identification number Dietary reference intakes Estimated average requirement Food frequency questionnaire Fibroblast like growth factor 23 Institute of medicine International units High performance liquid chromatography-tandem mass spectrometry Parathyroid hormone Recommended dietary allowance Standard deviation Socioeconomic status Solid phase extraction Sun protection factor Tolerable upper level of intake United states department of agriculture Ultraviolet

xi

Acknowledgements I owe my deepest gratitude to my supervisor, Dr. Sheila Innis, for giving me this opportunity and for the invaluable assistance, direction, support and time she has spent with me over the last two years, without which I would have been unable to complete this degree. I would also like to thank my supervisory committee members, Dr. David Kitts and Dr. Tim Green for their support and guidance throughout my graduate studies. I would like to acknowledge my fellow graduate students at the Innis lab, especially Kelly Mulder for her support and assistance throughout the program, and the tremendous amount of work she has dedicated to this project, this would not have been possible without her. I also would like to acknowledge Alejandra Wiedeman for her academic and emotional support. I would also like to thank Roger Dyer for his assistance with the laboratory analyses as well as his patience and diligence in answering even the most insignificant questions. Lastly, I would like to thank my family for their unconditional love and support.

xii

Chapter 1: Literature review 1.1

Introduction Vitamin D is a fat soluble vitamin responsible for maintaining calcium and phosphorous

homeostasis in the body, and is important for bone health. Recent interest in non-skeletal functions

of vitamin D has arisen following the discovery of proteins (enzymes and receptors) related to vitamin D metabolism in several tissues throughout the body not involved in bone metabolism. Thus, interest in vitamin D’s function beyond bone health has increased 1,2. New vitamin D intake recommendations were published in 2011 by the Institute of Medicine (IOM). These recommendations included cut-off values for serum (or plasma) 25 hydroxy vitamin D (25 (OH)D), as a biochemical marker of vitamin D deficiency, insufficiency and sufficiency. The IOM (2011) evaluated that, at the time of the development of the vitamin D intake recommendations, there was not enough evidence to support the development of dietary recommendations based on health outcomes other than bone 3. Previous studies have indicated that vitamin D intakes of Canadian children are low, despite fortification of milk and margarine with vitamin D 4. Few studies on vitamin D status in children below the age of 6 years were available at the time of the dietary development of the vitamin D recommendations, thus the cutoff values for 25 (OH)D and marker of vitamin D status set by the IOM (2011) were based mainly on studies using adults and adolescents. Currently, little information is available on whether vitamin D intakes below the recommendations in young children do result in biochemical evidence of vitamin insufficiency or deficiency. Some studies have suggested that the recommended vitamin D intakes for young children may be set too high, as few children with vitamin D intakes below the EAR appear to be vitamin D insufficient 5,6. Thus the main objectives of this thesis were to estimate vitamin D intake and the major dietary sources of vitamin D in young children in Vancouver. In addition, we sought to assess vitamin D status and determine the influence of low vitamin D intake, and the effect of season on 1

biochemical measures of vitamin D sufficiency, insufficiency and deficiency. The following chapter will give a brief review of vitamin D sources, metabolism of vitamin D and its physiological roles. This is followed by a review of the criteria used to define vitamin D sufficiency, insufficiency and deficiency. This chapter will then provide a review of current knowledge on vitamin D intake and vitamin D status in Canadian and US children below the age of 6 years.

1.2

Vitamin D sources Vitamin D is a fat soluble vitamin, and is considered a steroid hormone. Vitamin D can

be obtained both from the diet and by cutaneous synthesis, and as a result vitamin D status is determined by both factors 7. Individuals living in regions of the world at high latitudes rely primarily on dietary sources of vitamin D to fulfill vitamin D requirements during winter months, due to limited opportunity for endogenous synthesis in the skin during these months8,9. Thus, obtaining vitamin D from dietary sources is considered important for maintaining adequate plasma vitamin D concentrations. In the following, sources of vitamin D, including natural sources, fortified foods and fortification practices, and supplemental vitamin D will be discussed followed by a brief overview of cutaneous synthesis of vitamin D.

1.2.1

Diet, supplements, fortification and cutaneous synthesis Vitamin D is provided in the diet mainly as cholecalciferol (Vitamin D3), which is of

animal origin. Vitamin D can also be found as ergosterol (Vitamin D2) which occurs in small amounts in plant foods1. Both forms of vitamin D are produced industrially and may be used in supplements or used for fortification of foods 8,10 . Naturally occurring sources of vitamin D, however, are few, and these include foods such as fatty fish, organ meats and egg yolk.

2

Food containing naturally occurring vitamin D appear to be eaten infrequently in many westernized countries including Canada 4,11,12. Hill et al (2012) reported that only 2 % (0.09 µg/day) of the daily vitamin D intake of 4025 Canadians was provided by eggs, and 10 % was provided by fish corresponding to 0.43 µg /day 12. Fatty fish and egg yolk are foods not frequently eaten by many individuals in westernized North American diets, perhaps stemming in part from health concerns over the high cholesterol content of egg yolks 13 and mercury contamination in fish 14,15. The relatively low intake of foods naturally rich in vitamin D in modern westernized diets suggests that obtaining adequate amounts of vitamin D without food fortification or supplements is difficult. Fortification of foods with vitamin D in Canada in the 1940’s and 1950’s was initiated due to a high incidence of rickets among children 16. However, fortification practices were not well regulated and parts of the population had intakes too high while others had intakes considered too low, thus the regulation of the early fortification was changed. This meant that evaporated milk and milk powder was allowed added vitamin D, but fluid milk was not. Subsequently, the occurrence of rickets began to increase and the Canadian government once again allowed fortification of fluid milk in 1965 16. Currently, milk and margarine in Canada must contain approximately 2.5 µg vitamin D per cup (250 mL) and 13.0 µg vitamin D/100 g, respectively 17 (Table 1.1). In addition, goats’ milk, plant based milk alternatives and calciumfortified orange juices may be fortified with vitamin D. Cheese and yogurt are permitted to be produced from vitamin D fortified milk, but vitamin D is prohibited from being added directly to the product 18.

3

Table 1.1 Vitamin D content in foods Food Fortified Milk

250 mL

Vitamin D µg/ serving 2.5

Fortified orange juice

250 mL

2.5

1.0

Fortified plant beverages

250 mL

2.5

1.0

10 g

1.3

13.0

Egg yolk

1

0.8

4.7

Whole egg

1

0.7

1.3

Beef liver, fried

75 g

0.9

1.2

Tuna, bluefin, cooked

75 g

5.5

7.3

Salmon Sockeye, canned/bone

75 g

13.9

18.6

Salmon, Atlantic, baked/broiled

75 g

5.1

6.8

Canned tuna, in water

75 g

0.9

1.2

Cod, Atlantic, baked/broiled

75 g

0.9

1.2

Fortified margarine

Serving size

Vitamin D µg/100g 1.0

Data is derived from the Canadian nutrient file (CNF) 19.

Data collected by the Canadian community health survey (CCHS) showed that fortified milk was the largest contributor to vitamin D intake in the Canadian population, providing 49.1 % of the daily intake of vitamin D. However, despite fortification of milk, the total dietary vitamin D intake of Canadians appears to be low, providing an average total vitamin D intake from foods of 5.8 ± 0.9 µg/day 4. Vitamin D can also be obtained from supplements, however, their consumption by only a segment of the population means that concerns for low vitamin D intakes remains 20.

4

In addition to obtaining vitamin D from the diet, vitamin D can be synthesized endogenously from 7–dehydrocholesterol which is a derivate of cholesterol, found in the skin 21. When exposed to ultra violet (UVB) light (~290 to 320 nm), the β-steroid ring of 7dehydrocholesterol is broken and thereby converted into previtamin D3. Subsequently previtamin D3 undergoes isomerization converting previtamin D3 into Vitamin D3 (Figure 1.1) 1,7,10,22. The industrial production of vitamin D2 resembles that of the cutaneous synthesis of vitamin D3. First the vitamin is exposed to irradiation whereupon isomerization occurs resulting in production of vitamin D2. Vitamin D2 and D3 differ only in their side chain (Figure 1.1); their metabolism which is the same, will be reviewed in the following section 1,23.

5

Figure 1.1 Production of vitamin D3 and D2

A Isomerization

UVB 7-dehydrocholesterol

Previtamin D 3

B

Isomerization

Irradiation Ergosterol

Vitamin D3

Previtamin D2

Vitamin D2

A) Conversion of 7-dehydrocholesterol to vitamin D3. B) Conversion of ergosterol by irradiation to vitamin D2. The B-ring of 7-dehydrocholesterol and ergosterol is broken and both form the pre-form of the vitamin and both subsequently undergo isomerization to form vitamin D3 and D2. Figure modified from Bikle (2009) 23.

6

1.3

Vitamin D metabolism Vitamin D obtained in the diet or synthesized in the skin are both transported to the liver,

where they are metabolized identically 1,7. The following section provides a brief review of absorption and transport of vitamin D obtained from the diet as well as vitamin D derived from endogenous synthesis. This is followed by a description of the hepatic and renal metabolism of vitamin D from the diet and endogenous synthesis. Finally, the physiological roles of vitamin D in bone health and emerging evidence of non-skeletal functions of vitamin D are discussed as background to the research in this thesis.

1.3.1

Absorption and transport Vitamin D is a fat-soluble vitamin, thus vitamin D in the diet is best absorbed in the

presence of fat. After ingestion, dietary vitamin D is incorporated into micelles in the intestine and subsequently it diffuses into the intestinal enterocytes. Within the enterocytes, the vitamin is incorporated into chylomicrons, which enter the lymphatic system and thence enters the blood 24. Vitamin D is transported in the blood either by chylomicrons to the liver and extra hepatic tissues or bound to the vitamin D binding protein 1,23. After synthesis in the skin, cholecalciferol diffuses into the blood where it also binds to vitamin D binding protein (DBP) for transport to the liver or extra hepatic tissues. Excess vitamin D3 can be stored in adipose tissue for months to years and released when little vitamin D is synthesized in the skin 1,7,21. In the liver, the metabolism of D3 and D2 obtained from the diet, released from storage in adipose tissue, or vitamin D synthesized in the skin follow the same path of hydroxylation 1.

1.3.2

Hydroxylation of vitamin D in the liver and kidney Vitamin D must undergo two hydroxylations before it becomes biologically active. In the

liver, cholecalciferol is converted into 25- hydroxyvitamin D (25 (OH)D) by 25- hydroxylase, 7

which is thought to be cytochrome P450 2R1 (CYP2R1). Most of the newly synthesized 25 (OH)D is secreted into the blood where it binds to vitamin D binding protein (DBP), and has a half-life of approximately 15 days to three weeks 1,25. For the vitamin to become active, 25 (OH)D must undergo a second hydroxylation, which is initiated if there is a decrease in the calcium concentration in the blood. This decrease is detected by calcium sensing proteins in the thyroid gland, which in response secretes parathyroid hormone (PTH). The increase in PTH leads to up-regulation of 1α-hydroxylase (CYP27B1) in the kidney which in turn leads to synthesis of 1,25 (OH)D also called calcitriol, by hydroxylation of 25 (OH)D. Calcitriol is thus under tight homeostatic control and has a short half-life of a few hours 1,7,10,25. Inactivation and excretion of vitamin D is believed to be initiated by 24-α hydroxylase (CYP24), which is upregulated by increased levels of calcitriol itself and fibroblast like growth factor 23 (FGF23) 23. Through a series of steps, CYP24 converts the vitamin D metabolites 25 (OH)D and 1,25 (OH)D by 24-hydroxylation 7,23. This leads to the formation of calcitroic acid as the end product, which is secreted with the aid of bile acids, ultimately being excreted in the feces 1,7,23.

1.3.3

Physiological roles of vitamin D Vitamin D is known to function through its actions on the intracellular vitamin D

receptors (VDR), which influences gene transcription and thus protein synthesis. The best known actions of the proteins synthesized in response to binding of vitamin D to the VDR are those involved in the regulation of serum calcium and phosphate homeostasis 1,7. The homeostatic control of calcium and phosphate is regulated by two counteracting hormones, parathyroid hormone (PTH) and calcitonin, which are further described in the following section. Additionally, numerous studies have reported evidence that vitamin D has several positive effects on non-skeletal health outcomes. An overview of these new emerging areas of vitamin D are briefly reviewed in this section. 8

1.3.3.1

Calcium and phosphate homeostasis and bone health The best known roles of calcium and phosphorous in the body are in bone mineralization.

However, calcium has numerous other essential functions, including blood clotting, nerve conduction, muscle contraction, enzyme regulation, and membrane permeability 1. Many of these roles are reliant on a stabile calcium concentration in the blood, thus the regulation of calcium is vital. The main role of vitamin D is to regulate and maintain the calcium blood concentrations 7. Schematic of vitamin D metabolism is shown in Figure 1.2.

Figure 1.2 Schematic of vitamin D metabolism

Vitamin D from sun exposure

Vitamin D from diet

25-OHD in blood

↑1,25 OHD

↑1,25 OHD

↑ Ca 2+ absorption

↑ PTH

Blood Ca2+

↑ PTH

↓PTH

↑1,25 OHD

↑Calcitonin

↑ Bone calcium resorption

↑ Bone calcification

↑ Blood Ca2+

Schematic illustration vitamin D’s metabolism and regulation of calcium homeostasis. Modified from Holick (2007)10.

9

As previously introduced, when blood concentrations of calcium decrease, the parathyroid gland detects the decline and secretes PTH. PTH in turn activates 1α-hydroxylase and 1,25 (OH)D is synthesized. Calcitriol and PTH exerts their functions on three main target tissues which are the intestine, kidneys and bone, ultimately leading to increased calcium or phosphate concentrations in the blood. Vitamin D can act either by influencing gene expression or as a steroid hormone 7. In the intestine, calcitriol functions to increase calcium or phosphate absorption, by affecting gene expression and thereby initiating transcription of specific proteins involved in calcium absorption. Briefly, gene expression is initiated by 1,25 (OH)D binding to a nuclear vitamin D receptor (VDR). This complex is thought to bind to retinoic acid X receptor to form a heterodimeric complex, which in turn can interact with vitamin D response elements (VDREs) found on specific DNA sequences, ultimately resulting in regulation of gene expression, and thereby either enhancing or inhibiting transcription 1,7,21. As an example, calbindin D9k is a calcium binding protein found in the intestine which is synthesized in response to 1,25 (OH)D. After synthesis, calbindin functions on the brushborder to facilitate calcium absorption from the intestine. Additionally, it is believed that 1,25 (OH)D induces expression of epithelial cell calcium channels called epithelial calcium transient receptor potential vanilloid-type family member 6 (TRPV6) which are found on the brush border membrane of the intestine. The TRPV6 receptor interacts with calbindin D9k to increase calcium absorption in the intestine. Further, when calcium concentrations in the blood decrease, 1,25 (OH)D and PTH appear to stimulate calcium reabsorption in the kidney, thus decreasing the excretion of calcium 26. Calcitriol and PTH also play a role in mobilization of calcium and phosphate from the bone. Calcitriol or PTH also facilitates the formation and activation of osteoclasts by interacting with osteoblasts which in turn induces expression of receptor activator of nuclear factor-kB ligand (RANKL). RANKL then interacts with immature osteoclasts which induce maturation of 10

the osteoclasts that are responsible for mobilization of calcium and phosphate from the bone, causing calcium levels to rise. During times of vitamin D deficiency, calcium absorption from the intestine is decreased and mobilization of calcium from the bone is increased 1,7,21. Conversely, high serum levels of calcium or phosphorous lead to down-regulation by the hormone calcitonin which is secreted by the thyroid gland. Calcitonin promotes bone mineralization and blocks calcium mobilization from the bone 1. Further, calcitriol is thought to suppress PTH production and secretion when interacting with its receptor (VDR). Fibroblast like growth factor 23 (FGF23) is another protein involved in calcium and phosphorous homeostasis. It is produced in the osteocytes and osteoblasts in response to elevated levels of calcitriol and phosphate. It acts on the parathyroid gland to decrease PTH production. Further, FGF23 reduces CYP27B1 expression and increases expression of renal CYP24, thus decreasing 1,25 (OH)D synthesis and increasing the production of inactive metabolites of vitamin D. Further, FGF23 decreases the reabsorption and increases excretion of phosphate in the kidney by decreasing the expression of the sodium-phosphate co-transporter 1,7,23.

1.3.3.2

Emerging roles of vitamin D Recent interest in non-skeletal functions of vitamin D has arisen following the discovery

of vitamin D receptors (VDR), vitamin D responsive elements (VDREs), vitamin D metabolites and enzymes related to vitamin D metabolism in several tissues throughout the body not involved in bone metabolism. These tissues include the immune system, brain and placenta 27–30. Briefly emerging evidence relating vitamin D to immune function has found VDRs are present in several different cells of the adaptive and innate immune system, of which many are able to locally synthesize 1,25 (OH)D. It is believed that 1,25 (OH)D enhances the activity of the innate immune system and decrease the adaptive immune system 31, with 1,25 (OH)D also shown to enhance phagocytosis and modulate activated T and B lymphocytes. Experimental 11

studies have also indicated that 1,25 (OH)D is present in the cerebrospinal fluid of humans and can cross the blood brain barrier. Further, CYP27B1 and CYP24A1 have found to be present in the fetal and adult brain, thus 1,25 (OH)D can be produced locally in these cells or be inactivated 28. However, little is as yet known about vitamin D and brain development or function in human. Similarly, other studies have found CYP27B1 is present in the human placenta which is also capable of synthesizing 1,25 (OH)D and 24,25 (OH)D. It is believed that 1,25 (OH) D promotes anti-bacterial and anti-inflammatory responses and thus may function in placental immune and inflammatory response 31.

1.4

Consequences of low vitamin D status During vitamin D deprivation, calcium absorption from the intestine and reabsorption in

the kidney is decreased. This in turn prompts blood PTH concentrations to remain elevated and calcium mobilization from the bone persists. Thus, chronic low vitamin D status can cause bone demineralization and poor mineralization. Ultimately, vitamin D deficiency can lead to rickets in children, which is characterized by continuous growth of cartilage but failure to mineralize the bone 8. Adequate vitamin D status is especially important for children to ensure proper mineralization of bones 21.

1.5

Factors affecting vitamin D status In addition to vitamin D intake, several factors can influence an individual’s vitamin D

status. These factors include physical individual factors and environmental factors, including lifestyle choices, which will be discussed next.

1.5.1

Individual factors affecting vitamin D status A number of individual factors can affect vitamin D status. Studies have found that obese

individuals have lower 25 (OH)D concentrations than individuals considered to be normal 12

weight. This is believed to be caused by sequestering of vitamin D in the adipose tissue, but little is known about the cause and effect 32. Additionally, it is recognized that skin pigmentation affects the cutaneous synthesis of vitamin D. Darker skin types have more melanin which is believed to act as a natural sunscreen, and thereby decrease the amount of vitamin D synthesized in the skin 9. Ageing also has an effect on cutaneous synthesis of vitamin D, as the amount of 7dehydrocholesterol in the skin decreases with age. Therefore, elderly individuals may need longer sun exposure than younger individuals to achieve the same effect 9.

1.5.2

Environmental and lifestyle factors affecting vitamin D status The cutaneous synthesis of vitamin D can also be affected by environmental factors as

well as lifestyle factors. At higher latitudes, the intensity of the UVB rays reaching the earth’s surface is reduced during winter months, due to the earth’s inclination. Consequently, cutaneous synthesis of vitamin D is greatly reduced or does not occur 9,22. Further, overcast days during the year and air pollution are also believed to decrease the amounts of vitamin D synthesized in the skin 33. Thus, the availability of adequate vitamin D from the diet becomes particularly important, especially in areas of the world located at higher latitudes such as Vancouver, which is located at latitude of 49°N 16ʹ. Vancouver also has few hours of sunshine during winter and fall months namely, 60, 85, 134,182, 231 and 229 hours of sunshine for the months of January June and 294, 268, 199, 125 64 and 56 h/month for the 6 months of July - December, respectively 34. Additionally, concerns over skin cancer have lead to an increased use of sunscreen, sun avoidance by coverage of the skin with clothing or staying indoors 9,35,36. Use of sunscreen with sun protection factor (SPF) 15 and higher is believed to block nearly all cutaneous synthesis of vitamin D if applied as recommended 2. Further, increasing indoor activities, work and school also reduce sun exposure. Canadian children have physical activity levels below the recommendations of 60 min/day 37,38, thus out-door activity in children may 13

also have decreased. In addition, studies suggest that time spent watching TV and playing computer games have increased in children, possibly also leading to decreased time spent out-doors 39.

1.6

Assessment of vitamin D status The best biomarker for vitamin D status is considered to be 25 (OH)D, as it is not under

homeostatic control and has a relatively long half-life of approximately 15 days. Circulating levels of 25 (OH)D reflects the sum of vitamin D synthesized cutaneously and that obtained from the diet and supplements. Conversely, calcitriol is under homeostatic regulation and has a short half-life which limits its usefulness as a biomarker for vitamin D status. PTH has also been considered as a biomarker of vitamin D status due to the inverse relationship between PTH and 25 (OH)D. During a state of vitamin D deficiency, PTH remains elevated as the intestinal absorption of calcium and renal calcium reabsorption is decreased, which means that mobilization of calcium from the bone is needed to maintain circulating calcium concentrations. According to Prentice et al (2008), there is a wide variation of PTH concentrations within and among individuals, and several factors such as demographics, physiological factors, and other dietary variables can affect PTH levels. As a result, defining a normal PTH concentration has been difficult.

1.6.1

Vitamin D sufficiency, insufficiency and deficiency The most recent dietary recommendations for vitamin D and calcium from the Institute of

Medicine (IOM) were published in 2011 3. The 2011 Dietary Reference Intakes (DRI) for vitamin D noted that 25 (OH)D cut-points specifying vitamin D status had not undergone a systematic evidence based development process, and that several different cut-off levels were used by different expert groups. The recommendations for vitamin D that the IOM (2011) 14

developed are based on bone health as the sole outcome and do not include any roles of vitamin D other than skeletal health outcome, as they evaluated that the evidence base for other health outcomes were inconclusive. The IOM (2011) considered that there is evidence to believe that an increased risk of rickets, impaired calcium absorption, and decreased bone mineral density (BMD) occurs at serum 25 (OH)D concentrations below approximately 30 nmol/L. They concluded that maximal calcium absorption occurs at 25 (OH)D of 30 to 50 nmol/L, and that there is no evidence to suggest further benefits for bone health when 25 (OH)D increases above 50 nmol/L. Based on this, it is assumed that vitamin D intakes which achieve circulating levels of 40 nmol/L 25(OH)D is equivalent to the median requirement of the population, and that intakes that achieve circulating levels of 50 nmol/L would cover the needs of 97.5 % of the population. Due to the difficulty in estimating the amounts of vitamin D an individual will obtain from sun exposure, the vitamin D intake recommendations proposed by the IOM (2011) were developed assuming minimal sun exposure 3.

Several experts disagree with the 25 (OH)D cut-offs for vitamin D sufficiency, insufficiency and deficiency suggested by the IOM (2011). Holick (2007) argued that 25 (OH)D concentrations of 75 nmol/L should be considered sufficient because PTH and 25 (OH)D are inversely correlated until PTH begins to plateau at a 25 (OH)D of approximately 75 - 100 nmol/L, and intestinal calcium absorption is significantly increased at between 50 and 80 nmol/L 40. Similarly, Vieth (2011) advocated defining vitamin D sufficiency at 25 (OH)D >75 nmol/L 41. The Canadian Paediatric Society (CPS) defines vitamin D sufficiency, insufficiency and deficiency based on similar considerations to Holick (2007) and Vieth (2011). The CPS suggested that bone mobilization, PTH production and intestinal calcium absorption is stabile at plasma 25 (OH)D between 75 nmol/L to 225 nmol/L 42.

15

As shown in Table 1.2, utilizing the Canadian Paediatric Society’s (2007) classifications will identify a higher prevalence of individuals with vitamin D insufficiency than when using the cut-offs recommended by the IOM (2011). The Endocrine Society has suggested different cutoff for 25 (OH)D in children aged 4 – 8 years. The Endocrine Society has suggested that plasma 25 (OH)D below 50 nmol/L should be classified as deficiency and plasma 25 (OH)D concentrations above 75 nmol/L should be considered sufficient, and the Endocrine Society also suggested a daily vitamin D intake requirement for children age 4 – 8 years of 15 – 25 µg/d. In this thesis, vitamin D status will be defined based on the cut-off levels established by the IOM (2011), although the discussion will include commentary on the proportion of children meeting the criteria for insufficiency using the Canadian Paediatric Society (2007) and Endocrine Society cut-off values.

Table 1.2 Biochemical indicators of vitamin D sufficiency, insufficiency and deficiency Definition

IOM

CPS

ES

Deficient

< 30

< 25

< 50

Insufficient

30 – 50

25 – 75

52.5 – 72.5

Sufficient/Optimal

50 – 125

75 – 225

75- 250

> 125

> 225

-

Potential adverse effects

Values are nmol/L 25 (OH)D. Institute of medicine (2011), Canadian Paediatric Society (2007), Endocrine Society (2011).

1.7

Vitamin D intake recommendations As introduced in section 1.5, the amounts of vitamin D synthesized in the skin can vary

considerably among individuals and within individuals at different times of the year. The IOM committee reviewed studies that investigated the dose-response relationship between total 16

vitamin D intake and plasma 25 (OH)D under conditions of no or minimal sun exposure. They concluded that an intake of 10 µg/day (400 IU) and 15µg/day (600 IU) on average achieved a 25 (OH)D of 59 nmol/L and 63 nmol/L, respectively. Due to uncertainties such as inter-study variances and comparability of results from studies using different 25 (OH)D assay methods, the committee estimated that intakes10 µg/day (400 IU) and 15 µg/day (600 IU) would achieve a 25 (OH)D of 40 nmol/L and 50 nmol/L respectively 3. The vitamin D intake recommendations for children aged 4 - 8 years were extrapolated from studies in older children and adolescents and are based on measures of bone health 3.

1.8

Current knowledge of vitamin D intake and status in children At the time that the research in this thesis was designed and started, there was a small

amount of information on vitamin D intake in children, but information on vitamin D status of groups of Canadian 5 - 6 year olds was limited. The following section provides a summary of current information on vitamin D in children from two national surveys in Canada and the NHANES survey in the U.S. Further, research studies addressing vitamin D intake or status in young children in Canada and the U.S will be reviewed.

1.8.1

Canadian community health measures survey cycle 2.2 - 2004 The 2004 Canadian community health measures survey cycle 2.2 (CCHS 2.2) was a

cross-sectional study that collected nutritional information from a nationally representative sample of 34,789 Canadians of whom 5655 were children aged 1 - 8 years. The dietary information collected was based on a single computer assisted 24-hour recall from 31,107 participants. A second recall was collected from 10,786 participants, which constituted about 30 % of the sample population 4. Based on the initial recall, Canadian children had a mean intake of 17

vitamin D in 2004 of 6.2 ± 0.1 µg/day and a median intake of 5.6 (IQR 4.1 – 7.5) µg/day 4. The 75th percentile of intake was 7.5 µg/day, indicating that at least 75 % of the children consumed an amount of vitamin D below the EAR of 10.0 µg/day, as defined by the IOM (2011). The main source of dietary vitamin D for this age group was fortified milk, which contributed 75 % of the vitamin D intake 4. When the intake from supplements and diet was combined, the mean and median intake of vitamin D was 9.5 ± 0.2 µg/day and 8.2 (IQR 5.4 – 13.2) µg/day in children aged 4 – 8 year old children, respectively 43. The CCHS 2.2 did not collect blood samples from the population and thus vitamin D status could not be assessed.

1.8.2

Canadian health measures survey 2007 - 2009 and 2009 - 2011 The Canadian health measures survey (CHMS 2007 – 2009) is a nationally representative

survey of 5306 Canadians in which blood samples were collected and vitamin D status was measured. However, no children under the age of 6 were included in CHMS (CHMS 2007 2009). The 2007 – 2009 national survey showed a mean 25 (OH)D of 75.0 nmol/L and indicated that at the time, 95.6 % of 6 - 11 year olds had plasma 25 (OH)D concentrations above 37.5 nmol/L and 48.6 % had 25 (OH)D above 75 nmol/L 44. Fourteen percent (14.1 %) of the children within this age group had 25 (OH)D concentrations below 50nmol/L 20. Comprehensive dietary collection was not part of the CHMS, but information such as frequency of consumption of different food groups was collected 20. The survey showed that 28.7 % of the children were taking a vitamin D containing supplement, however, no information on frequency of the consumption or dose was available 20. Recently, some data from the CHMS Cycle 2 (2009 – 2011) was published 45. This cycle included children aged 3 – 5 years which the previous cycle had not. Of the 3 – 5 year old children, 518 children provided a blood sample46 and the mean plasma 25 (OH)D reported for the 3 - 5 year olds was 73.9 nmol/L and it was estimated that 11.0 % of the 3 – 5 year olds had a 25 (OH)D below 50 nmol/L. Of the 6 – 11 year olds, in the CHMS 18

(2009 – 2011), 974 provided a blood sample and 24.0 % of the children were found to have a plasma 25 (OH)D below 50 nmol/L. The mean 25 (OH)D was 67.3 nmol/L, and boys had a higher 25 (OH)D than girls of 72.0 nmol/L versus 63.0 nmol/L, respectively 45,47.

1.8.3

USA – National health and nutrition examination survey The National Health and Nutrition Examination Survey (NHANES 2005 - 2006) is a

national representative sample of the U.S population, which has been ongoing since the 1960’s and several series of this national population study have been conducted 48. The NHANES 2005 - 2006 estimated dietary and supplemental intake of vitamin D for the U.S population using two 24 hour recalls and a questionnaire, respectively. The available data published on vitamin D intake and status from the NHANES is summarized here. The 2005 - 2006 NHANES reported a mean vitamin D intake from diet alone for boys and girls 4 – 8 year olds of 6.4 ± 0.3 and 5.5 ± 0.3 µg/day, respectively. When including intake from supplements, the mean total intake of vitamin D for boys and girls aged 4 - 8 years increased to 9.3 ± 0.4 and 7.9 ± 0.6 µg/day, respectively. Supplements were given to 43 % of the boys and 34 % of the girls aged 4 – 8 years. The mean daily intake from supplements in children taking supplements was 6.6 ± 0.4 µg/day for boys and 7.9 ± 1.3 µg/day for girls49. Vitamin D status was reported for the NHANES 2001 - 2006 and this included 904 boys and 895 girls aged 1 - 5 years. The mean 25 (OH)D concentration was 71.0 nmol/L in boys and 70.0 nmol/L in girls. Of the total sample, 14 % was reported to have 25 (OH)D below 50 nmol/L and 63 % below 75 nmol/L 50. This estimate of 14 % of U.S children 1 – 5 years of age in 2001 – 2006 with a 25 (OH)D < 50nmol/L is similar to the estimate of 11 % of Canadian children 3 – 5 years of age in the 2009 – 2011 CHMS with a 25 (OH)D < 50 nmol/L 47.

19

1.8.4

Research reports At the time the research in this thesis was designed in 2010, little information had been

published on the vitamin D status of groups of Canadian children under the age of 6 years. The following section provides a review of recent studies on vitamin D intake and status of children in Canada and the U.S, other than those using national survey data.

1.8.4.1

Canada In 2005, Roth et al measured 25 (OH)D in 68 children seen in April 2003 in Edmonton,

which is at latitude 52°N. Of the 68 children aged 2 - 16 years, 35 of them were aged between 2 and 8 years. The mean 25 (OH)D for the 2 – 8 year olds was 51.5 nmol/L with an SD of 14.6 nmol/L. Of the 2 - 8 year olds, 2.9 % had a 25(OH)D below 25 nmol/L, but this represented only one child. However, 17 % (n= 6) had 25 (OH)D concentrations below 40 nmol/L, although the sample size of 35 children was small. For the entire group of 68 children, 34 % (n= 23) had a 25 (OH)D below 40 nmol/L and 5.9 % (n= 4) had a 25 (OH)D below 25 nmol/L. Of the entire group, 27 % were reportedly using a multivitamin regularly. The median vitamin D intake in the 2 – 8 years olds was 8.3 (IQR 4.6 – 10.4) µg/day with a lower intake of 5.8 (IQR 2.4 – 8.2) µg/day estimated using a FFQ and single 24 hr recall, respectively51. Hayek et al (2010) conducted a study including 388 Inuit preschoolers age 3 – 5 years living at latitudes between 51°N – 70 °N. They showed a daily mean vitamin D intake of 6.6 ± 2.9 µg/day (n= 279). Supplement use was recorded, as well as the frequency of supplement use and this showed that 3.7 % of the children were given a vitamin D supplement (presumably alone) and 16.8 % were given a multivitamin. Blood was collected from 282 children in summer and from 52 different children in winter (February – April). The median 25 (OH)D during summer was 48.3 (IQR 32.8 – 71.3) nmol/L. The children (n= 52) who were included in the study during winter had a median 25 (OH)D of 37.8 (IQR 21.5 – 52.0) nmol/L. This 20

corresponded to 51.7 % and 72.8 % of the children having 25 (OH)D below 50 nmol/L in summer and winter, respectively. In addition, vitamin D insufficiency was assessed using the Canadian Paediatric Society’s cut-off of (< 75.0 nmol/L), which classified 78.6 % and 96.8 % of children as insufficient during summer and winter, respectively 52. A more recent study by Hayek et al (2013) conducted between June 2010 and June 2011, assessed vitamin D status and intake in 508 children 2 - 5 year of age in Montreal, which is at latitude 45°N. They found that 88 % of the children had a 25 (OH)D ≥ 50 nmol/L and 49.4 % of the children had plasma 25 (OH)D ≥ 75nmol/L. However, 95 % of the children had vitamin D intakes below the EAR of 10 µg/day. The median dietary vitamin D intake was 5.9 (IQR 3.8 – 8.0) µg/day based on a single 24 hr recall. Approximately 28 % of the children were reportedly given supplements, estimated using a FFQ, with a median dose of 7.1 (IQR 3.2 – 10.0) µg/d in children given supplements. The total vitamin D intake from diet and supplements in all children was 9.9 (IQR – 7.1 -13.2) µg/day based on the FFQ 6. Another recent Canadian study had similar findings. Maguire et al. (2013) examined vitamin D status in 1311 children aged 2 – 5 years from Toronto, which is at latitude 43.4°N. The mean 25 (OH)D was 88 nmol/L; 35 % had a 25 (OH)D below 75 nmol/L and 6 % had concentrations below 50 nmol/L. The study did not collect comprehensive dietary data on vitamin D intake, but did collect data on frequency of consumption of cups of milk, and frequency of use of supplements. Of the 1311 children, 61 % reported taking supplements. The mean intake of cows’ milk was 455 mL/day, which would provide 4.7 µg/day vitamin D from milk. The authors reported a 6.5 % increase in 25 (OH)D for each one cup of milk consumed. Further, an increase of 13.2 % 25 (OH)D was attributed to daily use of vitamin D supplements. Winter was reported to result in decrease in 25 (OH)D of 10.7 % nmol/L when compared to summer. Skin pigmentation was also a major factor with a 9.9 % nmol/L difference between skin type I-III and IV -VI on the Fitzpatrick scale 53. 21

Hill et al (2012) conducted a study on vitamin D intake and the most common food sources of vitamin D among 4025 Canadians, of which 534 were children aged 2 – 12 years. They showed a mean intake of 4.4 ± 0.1 µg/day from food in the 2 - 12 year olds. No data on supplement use was included and data on proportion of vitamin D provided by different food sources was only reported for the whole study population, and not by age group 12.

1.8.4.2

U.S.A Abrams et al (2012) conducted a double blind randomized controlled trial investigating

whether supplementation with 1000 IU/day (25 µg/day) would affect calcium absorption in 4 - 8 year old girls (n= 64). The study was conducted in Texas which is located at latitude 29°N. The baseline vitamin D intake was 5.5 ± 0.2 µg/day which did not include supplemental vitamin D, as supplement use was an exclusion criteria. The mean ± SD baseline 25 (OH)D was 69.1 ± 18.5 nmol/L and no differences in baseline characteristics were found between the randomized groups. After 8 weeks of supplementation, the 25 (OH)D in the supplement group (n= 32) had increased to 89.9 ± 25.7 nmol/L. PTH had decreased from 21.4 ± 10.4 to 12.9 ± 7.1 pg/mL, with no significant changes from baseline observed in the placebo group (n= 31). Stable isotopes were used to assess calcium absorption, however, despite the increased 25 (OH)D and decreased PTH in the supplement group, no significant effects on calcium absorption was observed. Carpenter et al (2012) recruited 776 children aged six months to three years from Connecticut which is at latitude 41°N. They showed a mean ± SD vitamin D intake from food and supplements of 6.2 ± 3.2 µg/day. Children given vitamin D containing supplements above 10 µg/day were excluded from the study, which may mean the results do not reflect the general population. The children were mainly Hispanic (64 %) and African American (23 %), and had a mean ± SD 25 (OH)D of 66.0 ± 22 nmol/L, of which 15 % and 5.8 % had a 25 (OH)D below 50 22

nmol/L and 40 nmol/L, respectively. Despite classifying 15 % as vitamin D insufficient, only 2.5 % of the children showed elevations of PTH and alkaline phosphatase (ALP). Hill et al (2012) reported vitamin D intakes for 1350 U.S children aged 2 – 12 years. For these U.S children, the mean ± SD intake was 4.4 ± 0.1 µg/day, which was similar to the intake data for Canadian children described in the same report. The sample was selected based on the Canadian and U.S. census statistics aiming to reflect the respective populations12. Kemp et al (2007) collected blood samples from 142, 1 - 8 year old African American and Hispanic children from Newark (latitude 40°N). This study based on the relationship between blood lead concentrations and 25(OH)D during winter and summer, and its relation to child age and race. Blood was collected from all children in both winter and in summer, with the aim of estimating the increase in 25 (OH)D from winter to summer. The children were divided into two age groups, 1 - 3 years and 4 - 8 years. The mean 25 (OH)D for children aged 1 – 3 was 83.2 ± 3.0 nmol/L during winter and 84.2 ± 2.7 nmol/L during summer. Children aged 4 - 8 years showed a larger seasonal difference in 25 (OH)D with concentrations of 63.0 ± 3.0 and 84.2 ± 2.7 nmol/L in winter and summer, respectively. Of the entire group of children, 12 % had 25 (OH)D below 40 nmol/L in winter and 0.7 % had 25 (OH)D below 40 nmol/L in summer 54. In 2008, Gordon et al recruited 380 infants and toddlers aged 8 - 24 months from Boston (latitude 42°N). The population was primarily African American or Hispanic. The mean 25 (OH)D for infants and toddlers was 87.9 ± 37.9 and 85.6 ± 30.7 nmol/L, respectively. Forty percent (40 %) had a plasma 25 (OH)D below 75 nmol/L, 12.1 % were classified as deficient (defined as ≤ 50 nmol/L) and 1.9 % had severe deficiency (defined as a 25 (OH)D below 20 nmol/L). Forty participants were classified as deficient, and returned for a radiographic assessment of the wrist and knee. It was estimated that 32.5 % of these children showed signs of demineralization and one child showed signs of rickets 55.

23

Stein et al (2006) conducted a cross-sectional study that included 168 girls aged 4 – 8 years to investigate the relationship between 25 (OH)D and bone area, bone mineral density (BMD) and bone mineral content (BMC). The study was conducted in Athens U.S (latitude 34°N) and included white (n= 120) and black (n= 48) girls who were found to have a mean ± SD plasma 25 (OH)D of 99.2 ± 28.2 nmol/L and 80.4 ± 23.1 nmol/L, respectively. Only four children had a plasma 25 (OH)D below 50 nmol/L corresponding to 2.4 % of the total sample. The authors concluded that 25 (OH)D concentrations were not positively associated with BMD or BMC 56.

1.8.5

Summary The available data published over about the last decade suggests that vitamin D intake in

many Canadian as well as U.S children below 6 years of age is below the current EAR of 10 µg/day and RDA of 15 µg/day (Table 1.4, Table 1.5). However, the limited data on vitamin D status based on measures of 25(OH)D indicates that the proportion of children with vitamin D insufficiency is much lower than the proportion of children not meeting the EAR (Table 1.6, Table 1.7). However, current data also suggest that many children are given vitamin D containing supplements, although the frequency of dose given varies considerably among different studies (Table 1.3).

24

Table 1.3 Summary of the proportions of children given supplements in different studies Age

Sex

Supplement use

CHMS (2007 - 2009)1

1–8

Both

28.7

Hayek (2010)2

3–5

Both

20.5

Hayek (2013)

2–5

Both

27.7

Maguire (2013)

2–5

Both

61.0

Roth (2005)3

2–8

Both

27.0

NHANES (2005 - 2006)

4–8

Boys

43.0

NHANES (2005 – 2006)

4–8

Girls

34.0

Stein (2006)4

4–8

Girls

61.0

Canada

USA

1

Data derived from Whiting et al (2011); 2 3.7 % Vitamin D containing supplement and 16.8 %

multivitamin, does not specify whether they contained vitamin D; 3MVI use, does not specify whether they were vitamin D containing; 4Data reported for Caucasian girls, supplement use of black girls 32 %.

25

Table 1.4 Summary of vitamin D intakes in children in Canada Reference

Sex

Age

n

Vitamin D intakes µg/d Mean ± SD

Median (IQR)

Source

Collection method

CCHS 2.2 (2004)

Both

1–8

5655

6.2 ± 0.1

5.6 (4.1 - 7.5)

Foods

Single 24 hr recall

CCHS 2.2 (2004)

Both

1–8

5655

9.5 ± 0.2

8.2 (5.4 - 13.2)

Food + supp

Two 24 hr recalls1

Hayek et al (2010)

Both

3–5

275

6.3 ± 2.9

N/R

Unclear2

Single 24 hr recall

Hayek et al (2013)

Both

2–5

479

N/R

5.9 (3.8 - 8.0)

Food

Single 24 hr recall

Hayek et al (2013)

Both

2–5

479

N/R

9.9 (7.1 - 13.2)

Food + supp

FFQ1

Hill et al (2012)

Both

2 – 12

534

4.4 ± 0.1

N/R

Food

7 – 14 diet records

Roth et al (2005)

Both

2–8

35

N/R

8.3 (4.6 – 10.4)

Food + supp

FFQ1

Roth et al (2005)

Both

2–8

25

N/R

5.8 (2.4 – 8.2)

Food

Single 24 hr recall

N/R – Not reported; 1Supplement data collected by questionnaire; 2Does not specify whether or not supplements have been included in the estimate.

26

Table 1.5 Summary of vitamin D intake in children in USA

Reference

Sex

Age (Yrs)

n

Vitamin D intake µg/d Source

Collection method

Mean ± SD NHANES (2005 – 2006)

Boys

4–8

431

6.4 ± 0.3

Foods

Two 24 hr recalls

NHANES (2005 – 2006)

Boys

4–8

431

9.3 ± 0.4

Food + supp

Two 24 hr recalls1

NHANES (2005 – 2006)

Girls

4–8

468

5.5 ± 0.3

Food

Two 24 hr recalls

NHANES (2005 – 2006)

Girls

4–8

468

7.9 ± 0.6

Food + supp

Two 24 hr recalls1

Abrams et al (2012)2

Both

4–8

62

5.5 ± 2.0

Food

Single 24 hr recall & three day diet record

Carpenter et al (2012)3

Both

6 mo - 3 yrs

776

6.2 ± 3.2

Food + supp

Three day diet record

Hill et al (2012)

Both

2 – 12

1350

4.4 ± 0.1

Food

7 – 14 days diet record

Stein et al (2006)4

Girls

4–8

114

9.7 ± 5.7

Food + supp

Three day diet record1

Median values were not reported for any of the studies; 1 Supplement data collected by questionnaire; 2 Study excluded children taking supplements; 3Study excluded children taking > 10 µg supplemental vitamin D; 4Results shown for white girls only, vitamin D intake among black girls (n= 42) were 6.8 ± 5.5 µg/day.

27

Table 1.6 Summary of vitamin D status of children in Canada Reference

Latitude

Sex

N°

Age

n

(Yrs)

Vitamin D nmol/L

< 50.0

≥ 75.0

nmol/L(%)

nmol/L(%)

CHMS (2007 - 2009)

> 42°

Both

6 – 11

903

75.0

14.1

48.6

CHMS (2009 - 2011)

>42°

Both

3–5

518

73.9

11.0

N/R

CHMS (2009 - 2011)

>42°

Both

6 - 11

974

67.3

24.0

N/R

CHMS (2009 - 2011)1

>42°

Boys

6 -11

N/R

72.0

N/R

N/R

CHMS (2009 - 2011)1

>42°

Girls

6 – 11

N/R

63.0

N/R

N/R

Hayek et al (2010)2

51° - 70°

Both

3–5

282

48.3 (32.7 – 1.4)

51.7

78.6

Hayek et al (2010)3

51° - 70°

Both

3–5

52

37.8 (21.5 – 2.9)

72.8

96.8

Hayek et al (2013)

45.0°

Both

2–5

508

74.4 (60.3 – 93.5)

10.6

49.4

Maguire et al (2013)

43.4°

Both

2–5

1311

88.0

6.0

35.0

Roth et al (2005)

52.0°

Both

2–8

35

51.5

N/A

N/A

N/R = not reported. Data in bold are median and data in italic are means; 1n was reported for children who provided a blood sample but the actual sample size used to estimate the mean plasma 25 (OH)D was not reported, should be interpreted with caution; 2Summer; 3Winter.

28

Table 1.7 Summary of vitamin D status of children in the U.S

Reference

Latitude

Sex

N°

Age

n

(Yrs)

Vitamin D

< 50.0

≥ 75.0

nmol/L1)

nmol/L (%)

nmol/L(%)

NHANES (2001 – 2006)

National

Boys

1–5

904

71.0

14.0

39.0

NHANES (2001 – 2006)

National

Girls

1-5

895

70.0

14.0

35.0

Abrams (2012)

29°

Both

4–8

62

69.1 ± 18.5

N/R

N/R

Carpenter (2012)

41°

Both

3 mo – 3y

781

66.0 ± 22.0

15.0

N/R

Gordon (2008)

42°

Both

8 - 24 mo

2472)

85.6 ± 30.7

10.8

40.4

Gordon (2008)

42°

Both

8 - 24 mo

1333)

89.9 ± 37.9

14.4

39.2

Kemp (2007)

40°

Both

1 – 34)

78

84.4 ± 2.8

N/R

N/R

Kemp (2007)

40°

Both

1 – 34)

78

83.4 ± 3.0

N/R

N/R

Kemp (2007)

40°

Both

4 – 85)

64

84.4 ± 2.8

N/R

N/R

Kemp (2007)

40°

Both

4 – 85)

64

63.2 ± 3.0

N/R

N/R

Stein et al (2006)

34°

Girls

4–8

168

93.8 ± 28.1

2.4

N/R

1)

Data is presented as means and included SD where indicated 2) Infant group – no specific definition of age 3) Toddler group – no

specific definition of age, 4)Summer, 5)Winter.

29

Chapter 2: Study 2.1

Purpose Despite the recognized importance of nutrition in young children and the known

importance of vitamin D in calcium homeostasis and bone mineralization, knowledge of vitamin D intakes and status among young children is limited. The most recent DRIs for vitamin D issued by the U.S Institute of Medicine in 2011 for children 4 - 8 years of age were based on limited scientific data regarding vitamin D status in this age group. Although not addressed in this thesis, plasma 25 (OH)D that best reflect vitamin D deficiency, insufficiency, and sufficiency in young children remain unclear, and to date no set cut-offs have been universally agreed upon by different expert groups. This study seeks to provide knowledge on dietary and supplemental intakes of vitamin D

and the effects on plasma vitamin D status, based on measures of plasma 25 (OH)D in children aged 5 - 6 years of age living in Vancouver, British Columbia. Vancouver is located at latitude 49°N and with an average of 60, 85, 134,182, 231 and 229 hours of sunshine for the months of January to June and 294, 268, 199, 125 64 and 56 h/month the months of July to December 34, and is a geographical area at which its inhabitants may be at risk for low capacity for endogenous vitamin D synthesis for much of the year. The purpose is to measure plasma 25 (OH)D concentrations, and assess concentrations in children compared to cut offs for deficiency, insufficiency, and sufficiency recommended in the 2011 DRI, and among children grouped by dietary intake of vitamin D, use of vitamin D supplements, and by season of the year.

2.2

Objectives The research in this thesis is designed with the following objectives for children 5 - 6

years of age living in Vancouver:

1. To estimate dietary vitamin D intake from natural and fortified food sources, and the intake of vitamin D from supplements 2. To determine the proportion of children meeting the EAR and RDA as set by the IOM (2011) based on their estimated vitamin D intake 3. To determine vitamin D status based on measures of plasma 25 (OH)D for whom dietary vitamin D has been estimated 4. To determine the proportion of children who, based on their plasma 25 (OH)D meet the criteria for vitamin D deficiency, insufficiency and sufficiency 5. To assess the importance of season of the year and vitamin D intake from foods and supplements in contributing to vitamin D deficiency and/or insufficiency when compared to vitamin D sufficiency

2.3 2.3.1

Methods Design and setting This study is a cross-sectional study which was conducted at the University of British

Columbia, Oak St Campus, at the Child and Family Research Institute. Children and their parent or legal guardian (abbreviated as parents), all residents of greater Vancouver, were enrolled from the community between July 2010 and March 2013. Parents interested in participating with their child were asked to attend a research clinic, when their child was between 5 years 8 months and 5 years 11 months of age. Prior to collection of any information or test, the parent signed an informed consent form (Appendix A). Participants were assigned a randomly generated 4-digit 31

subject code and this was used on all data collection forms to maintain anonymity during the study. The range of ages over which children were seen was relatively narrow because of the potential for changes in developmental skills as well as food likes/dislikes with increased exposure to school, which cannot be adequately adjusted for by statistical approaches to control for differences in ages. Ethics approval for this research project was obtained from the University of British Columbia / Children’s and Women’s Health Centre of British Columbia Research Ethics Board (UBC C&W REB) a UBC-affiliated Research Ethics Board (REB) for the Oak Street campus. In addition, ethical approval to approach and recruit subjects in the community was obtained from Vancouver Coastal Health, and Vancouver School Board Research Ethics Committee.

2.3.2

Inclusion and exclusion criteria and recruitment Inclusion criteria were that the parent (usually the mother) was comfortable speaking and

writing English language, and that the child was born after full-term gestation (>37 wk gestation) with no congenital or acquired disease considered likely to impact healthy child growth and development. The children had to be between the ages of 5 years 8 months and 5 years 11 months of age at the time when they participated in the study. Children with iron deficiency were to be withdrawn from the study, based on hematocrit levels below 34. Exclusion criteria were all children not meeting the inclusion criteria.

2.3.3

Demographic characteristics Information on each child’s family background was collected using confidential

questionnaires labeled only with the child’s random number code. Information on the mother’s age, ethnic background, highest level of education, number of adults and children in the

32

household, smoking and other relevant information was collected. Total annual family income was recorded as $20,000 (or less) to $80,000 (or more) in $10,000 increments (Appendix B).

2.3.4

Dietary assessments and collection of information on supplement use To assess each child’s dietary intake, both a Food Frequency Questionnaire (FFQ) and

three 24-hour dietary recalls were conducted. The FFQ interview and first 24 hour dietary recall were conducted by myself or another trained interviewer during the study visit. The 24 hour recall was done with food models, cups, spoons and packages to assist the parent in estimating portion sizes. A multiple step approach (5 pass) was used for the 24 hour recalls 57,58. Briefly, the participant was asked to list uninterruptedly, all foods and beverages consumed by the child during the previous 24 hours. Next, the participant was asked about common forgotten foods, such as condiments, drinks and snacks. The participant was next asked for time, location and occasion when the foods and beverages were consumed. Using the food models, the parent was then asked for descriptions and quantities of foods and beverages consumed. Following this, the list was reviewed to ensure no foods had been forgotten. Finally, the participant was asked if the child had been given any supplements. If so, brand and amount of supplements given were recorded on the bottom of the page. If no supplements had been given “No supplements“ was recorded to indicate that the participants had been asked this. A second and third unscheduled 24- hour recall was conducted by telephone using the exact same procedure. Each parent was asked for the most convenient time to call. The random telephone calls were conducted over the following two weeks to capture a total of one weekend day and two week days of records for each child. The FFQ was interviewer-administered, and utilized to collect details on the frequency and portions sizes of foods eaten over the past 4 weeks. The FFQ was used to capture details on the intake of foods, in particular major protein sources, such as fish and meats that are not eaten 33

on a daily basis and for which single or 3 day records may therefore result in errors in estimating average intakes for the individual. The FFQ contains 14 main food categories that include different food items, with sub-categories for fat content; for example whole, 2 %, 1 % or skim milk. Another example is the specific types of fats, such as margarine and butter, yogurts, cream cheeses, types of vegetable oils, salad dressings and mayonnaise, fresh, frozen and canned fish, and types of fish and shell fish, as well as eggs, meats and poultry. The FFQ collects information on both the amount and frequency of consumption, and allows for addition of items not on the list, or other portion sizes (Appendix C). Further, information on supplement use, including brands and when the use of these commenced was collected with the FFQ.

2.3.5

Analysis of dietary intakes The dietary information was entered into nutrient analysis software (ESHA Food

Processor SQL. Version 10.10.0.0, Salem, OR: ESHA Research, 2012), to enable analysis of each child’s intake of energy and all nutrients, and to create food lists with the amounts of different nutrients from different foods. The nutrient software contains the Canadian Nutrient File (CNF) 19 and United States Department of Agriculture (USDA) nutrient data base for several thousand foods and specific food brands. Home prepared foods and restaurant foods, as needed, were disaggregated for entry into the database. The FFQ and 24 hour recalls were entered by myself or another research assistant, and vitamin D intakes retrieved from the database as µg/day. All data was cross-checked to ensure accuracy and consistency of the data entry. Dietary vitamin D intake was calculated using both the FFQ and three 24 hr recalls and intakes were calculated as the average intake in µg/day and intake/1,000 calories/day. Supplement data was calculated using both the FFQ and three 24 hour recalls, and intakes 34

calculated as the average intake/day. When the supplement brand or amounts of supplements given was not recorded or known by the participant a standard amount of 10 µg vitamin D was assumed or the intake recommendations of the manufacturer of a specific brand used when type but not dose was known. Incomplete data was entered as missing values.

2.3.6

Anthropometrics For each child, standing height and weight was measured using standardized, calibrated

equipment in the Clinical Research Unit at the Child and Family Research Institute. Height and weight were each measured twice and reported as an average of the two. Z-scores for height-forage, weight-for-age and BMI-for- age were calculated using the World Health Organization’s (WHO) AnthroPlus software 59.

2.3.7 Blood collection, preparation and analysis of 25(OH)D Venous blood was collected by a registered technician following completion of all other assessments, and data collection with the child and parent. Directly after collection, the blood was transferred to the laboratory, then immediately centrifuged at 3700 rpm for 10 min, 4°C to separate the plasma and blood cells. The plasma was aliquoted into storage tubes, sealed, labeled and stored at -80 °C until analysis.

2.3.7.1

LC-MS/MS Plasma concentrations of (25OH)D were used as a biochemical measure for vitamin D

status using LC- tandem mass spectrometry (MS/MS). The LC-MS/MS is a Waters ACQUITY UHPLC system connected to a Quattro micro triple quadrupole mass spectrometer (Waters Canada, Mississauga, Ontario). Briefly, deuterium-labeled 25-hydroxy vitamin D3-d6 Calcidiol 35

(26,26,26,27,27,27-d6) (Chemphor, Canada, Ottawa, Ontario # CHE011) in 15 µl was added to 75 μL plasma. Proteins were precipitated with 1 volume of acetonitrile, the mixture vortexed for 10 seconds, then left at room temperature for 5 minutes. Thereafter, the mixture was vortexed again for 5 - 10 seconds, and then centrifuged at 15,000g for 10 minutes to pellet the proteins and the supernatant was recovered. Next, 400 µl ammonium formate (in 0.01 M water) was added, the sample vortexed 5-10 seconds, 1ml of ethyl acetate, added the tube vortexed for 10 seconds and then left at room temperature for 2 - 3 minutes. After vortexing again for 5 - 10 seconds, the sample was centrifuged at 2000 rpm for 5 minutes to obtain well defined aqueous and organic layers. The top layer (ethyl acetate) was removed to brown tinted autovials and dried under nitrogen. An additional 1ml of ethyl acetate was added to the remaining lower layer to reextract and recover any 25 (OH)D remaining, and this was pooled with the first extract. The samples were then dried under nitrogen, 75µl of the derivatizing agent, 1 mg/ml 4-phenyl-1,2,4triazonile-3,5-dione (Sigma-Aldrich, Canada, Oakville, Ontario, # 280992) in acetonitrile (PTAD/acetonitrile) was added and the vials left in the dark for 1 hour at room temperature. Then, 55 µl ammonium formate (in 0.01 M water) was added and the samples were transferred to 150 µl autosampler inserts each with a clean plastic transfer pipette. Chromatography was accomplished on a ACQUITY UPLC BEH C8 column, 1.7µm, 2.1mm x 50 mm (Waters Canada, Mississauga, Ontario) with a binary mobile phase gradient which were, ammonium formate in water (A) and methanol (B), both containing 0.1 % formic acid (v/v), commencing at 85 % (A) 15 % (B) and changing to 30 % (A) and 70 % (B). The column was flushed with 100 % methanol before returning to initial conditions. The sample injection volume was 3 μL and accomplished with an auto-sampler and temperature-controlled sample chamber held at 5oC. A standard curve was created using standard solutions of 5, 10, 20, 40, 60, 90, 120, 150 and 200 ng/mL. Quality control samples were made using internal standard (15ul) plus 75µl of methanol, 36

with a reagent blank made using 90 µl methanol were analyzed with each sample batch. The intra assay CV was 3.6 % and the inter-assay was 8.7 %. In method development, serum samples in which 25 (OH)D had been analyzed by radio- immunoassay were generously provided by Dr. Tim Green, University of British Columbia, and these were analyzed to compare the results as analyzed by LC-MS/MS. The correlation coefficient between the RIA and UHPLC analyses was r= 0.88.

2.3.8

Statistical analysis Statistical analyses were performed using IBM SPSS statistics version SPSS software

(IBM SPSS Statistics for Windows, Version 21.0, 2012. Armonk, NY: IBM Corp). All dietary data was checked for normality using the Kolmogorov-Smirnov’s test of normality, with a significance level < 0.05 used to indicate a deviation from the normal distribution. Descriptive statistics, including, percentages, means ± SD, medians and inter quartile ranges (IQR), and 2.5th – 97.5th percentile were used to summarize the data collected. Subsequently, the statistical analyses were based on means and medians using parametric and non-parametric tests as applicable. Pearson correlation was used to determine the correlation within individuals between energy intake as well as vitamin D intake when assessed using the FFQ compared to a single 24 hr recall, or the average of three 24 hr recalls. The proportion of dietary vitamin D obtained from different foods (contribution of major dietary sources to vitamin D intake) was calculated for each child using data from the FFQ. Additionally, the proportion of vitamin D derived from natural, fortified foods and supplements were calculated for each child, then the group means and medians calculated. Children were grouped based on vitamin D status as sufficient (≥ 50 nmol/L), insufficient (50 – 30 nmol/L) and deficient (< 30 nmol/L) based on the cut-off values defined by the IOM (2011) 3.The proportion of children who were deficient, insufficient and sufficient during different seasons was also determined. The FFQ and three 24 hr recalls were 37

used to quantify the proportion of children meeting the EAR and RDA for vitamin D based on diet alone, as well as diet and supplements in combination. Further, children were split into supplement users and non-supplement users, and the proportion of children meeting the EAR and RDA for each group determined for the two dietary data collection methods was quantified. Spearman correlation was used to determine whether there was a correlation between vitamin D intake and plasma 25 (OH)D. A P value ≤ 0.05 was considered significant.

38

Chapter 3: Results Figure 3.1 Study flow chart Consented, n=200

Research study appointment1

Child assessment

Parental interviews

Demographics, n=1942

Anthropometrics, n=199

FFQ, n=1963

Blood sample, n=154 n 4

1 x 24 hr recall, n= 194

3 x 24 hr recall, n= 1835

Supplement data FFQ, n=1936 Supplement data 3 x 24 hr recall, n=181

1

Excluded, n= 1 did not meet inclusion criteria (second of a twin pair); 2 demographic data

excluded, n= 5 (they were fathers); missing data for ethnicity and household income, n=1; excluded due to incomplete data, n=3; 4 excluded sick child or incomplete data, n= 5; 5 lost to follow-up or unreliable data for 24 hr recalls, n=9; 6supplement data n= 81 assumed manufacturer recommended dose, and n =46 assumed content (10 µg for MVI).

39

3.1

Subject characteristics From July 2010 to March 2013, 200 children who met all inclusion criteria were enrolled

in the study. However, one twin pair was enrolled, and therefore the second child of the twin pair was excluded, as per protocol. No child was found to have a low hematocrit, thus none were excluded on the basis of concerns over this measure of iron status. Of the children, 48.2 % were boys and 51.8 % were girls. The children’s age was controlled by the study inclusion criteria, and thus showed little variation, with a mean age of 68.7 ± 0.7 months. Of the parents who attended the research study, 194 were mothers and five were fathers. Information on household income was reported by n= 193 parents of whom a large proportion reported an annual household income above $50,000 CAD (85.0 %). An income between $30,000 to $50,000 CAD was reported by 9.8 % of parents and 5.2 % reported having an annual household income below $30,000 CAD. Information on ethnic background for both mother and father were recorded, but only information for the mother is presented. Information on the mothers’ ethnicity was available for n= 193, of whom 70.0 % were of Caucasian background, 14.0 % were Chinese, and 3.6% were of East Indian background. Further, 3.6 % were of other Asian backgrounds, and 9.3 % were grouped as from other backgrounds as their mothers identified themselves having a mix of ethnicities. No individuals of African American background and only one individual of First Nations descent (corresponding to 0.5 %) were enrolled in the study, with the latter child grouped under the “others” category. In addition, of the parents from whom information on education was available (n=194), 28.9 % reported having a university graduate degree, and 39.7 % reported having a university undergraduate degree. Further, 26.8 % reported having a college degree or a diploma and 4.6 % reported having a high school diploma or less. Maternal age at the research study visit was divided into four groups; 2.6 % reported being in the age category 20 –

40

29 years, 43.8 % were between 30 – 39 years, 53.6 % were 40 – > 49 years, with one woman (0.5 %) being over fifty years (Table 3.1).

41

Table 3.1 Subject characteristics Characteristic

N

Sex of children (%), boys / girls

199

48.2 / 51.8

Age in months at time of visit (Mean ± SD)

199

68.7 ± 0.7

No. of children in home (Mean ± SD)

198

2.0 ± 0.6

No. of individuals in home (Mean ± SD)

198

4.1 ± 0.9

Household income level (%)

193

< $30,000

5.2

$30,000 - $50,000

9.8

> $50,000

85.0

Mother’s ethnicity (%)

193

White/Caucasian

70.0

Chinese

14.0

Other

9.8

East Indian

3.6

Other Asian

3.6

Mother’s highest education level (%)

194

High school or less

4.6

College/diploma

26.8

University undergraduate degree

39.7

University graduate degree

28.9

Mother’s age (%)

194

20 – 29 years

2.6

30 – 39 years

43.8

40 - > 49 years

53. 6

Of the children included in the study n= 96 were boys and n=103 were girls. 42

Based on the World Health Organization’s (WHO) reference curves59, the mean ± SD zscore for weight-for-age was 0.18 ± 0.95, and 75.4 % had weight-for-age within -2 SD to +2 SD, 7 % were classified as low weight-for-age ( +2 SD, including one child > +3 SD). The mean height-for-age z-score was 0.17 ± 0.97, and 70.4 % were within the normal range, 9.0 % had a low height-for-age and 20.6 % had a height-forage z-score > +2SD. The mean BMI-for-age was 0.09 ± 0.99, 9.5 % had a low BMI z-score +2 SD, and 75.9 % had a BMI-for-age between -2 SD to +2 SD. Only one child had a BMI-for-age z-scores (and weight-for-age z-score) > +3 SD and this child was included in the previous category (Table 3.2).

Table 3.2 Child anthropometric measures

Weigh-for-age Z-score

0.18 ± 0.95

Height-for-age Z-score

0.17 ± 0.97

BMI-for-age Z-score

0.09 ± 0.99

Low BMI-for-age (%)

9.5

Normal BMI-for-age (%)

75.9

High BMI-for-age (%)

14.6

Results are as means ± SD, % as indicated, n=199. Low BMI-for-age was defined as the percentage of children with a BMI Z-score +2 SD, based on WHO’s growth curves. Only one child had BMI-for-age >3+SD as was included in the high BMI-for-age category.

43

3.2

Vitamin D intake Dietary vitamin D intake data was skewed to the right, with a higher mean than median

when estimated using both the FFQ and 24 hr recalls. Analysis of the FFQ, the single 24 hr recall for the day preceding the study visit, and the average of three 24 hr recalls all showed that boys had a higher energy intake than girls. Boys and girls, respectively, had median energy intakes of 1833 (IQR 786) kcal/day and 1690 (IQR 545) (p= 0.019) estimated using the FFQ, 1651(IQR 612) kcal/day and 1344 (IQR 548) kcal/day (p< 0.001), respectively, using a single 24 hr recall, and 1577 (IQR 452) kcal/day and 1388 (395)kcal/day, respectively, using the three 24 hr recalls (p