Anatomical Aspects of the Development of the Sense of Taste in Humans

Maryam Correa (nee Shahbake) BSc (Hons I)

This thesis was submitted in fulfilment of the requirements for the degree of Doctor of Philosophy to University of Western Sydney, College of Health and Science, School of Natural Sciences.

September 2016 Content

Declaration VI Acknowledgements VII List of Figures VIII List of Tables XIV List of Abbreviations XV Abstract XVII

Chapter 1: General Introduction 1

1.1 Anatomy of the Peripheral Gustatory System 4 1.1.1 The Tongue 4 1.1.2 Taste Papillae 5 1.1.2.1 Fungiform Papilla 5 1.1.2.2 Circumvallate Papillae 7 1.1.2.3 Foliate Papilla 7 1.1.2.4 Conical Papilla 8 1.1.2.5 Filiform Papillae 9 1.1.3 Taste Buds 9 1.1.3.1 Taste Bud Distribution 11 1.1.3.2 Taste Bud Innervation 12 1.2 Development of the Peripheral Gustatory System 14 1.2.1 Foetal and Neonatal Development 14 1.2.1.1 Animal Studies 14 1.2.1.2 Human Studies 17 1.2.2 Infant to Adult Development 19 1.2.2.1 Animal Studies 19 1.2.2.2 Human Studies 20 1.2.3 Adult to Elderly 21 1.3 Methods Used to Study Peripheral Gustatory Anatomy 23 1.3.1 Taste Papillae 23 1.3.1.1 Magnifying Glass 23 1.3.1.2 Light Microscopy 24 1.3.1.3 Video-microscopy 25 1.3.1.4 Cell Counters 28 1.3.1.5 Contact Endoscopy 29 1.3.1.6 Summary of Findings Using Different Methods 34 1.3.2 Taste Buds 35 1.3.2.1 Light Microscopy 35 1.3.2.2 Electron Microscopy (EM) and Scanning Electron Microscopy (SEM) 36 1.3.2.3 Video-microscopy 37 1.3.2.4 Summary of Findings Using Different Methods 38 1.4 Behavioural and Psychophysical Studies of the Development of the Gustatory System 40

I 1.4.1 Newborn to Infant 40 1.4.2 Adolescent to Adult 41 1.4.3 Gustatory Perception 41 1.5 Diseases and Taste Disorders 44 1.5.1 Taste and Common Diseases in Children 46 1.5.1.1 Cystic Fibrosis 46 1.5.1.2 Diabetes Mellitus 47 1.5.1.3 Otitis Media 47 1.5.1.4 Radiation Therapy 48 1.5.1.5 Renal Disease 49 1.6 Summary 50 1.7 Aims of Thesis 51

Chapter 2: Development of a Rapid Method for Counting Fungiform Papillae 52

2.1 Introduction 53 2.2 Aims 54 2.3 Methods 54 2.3.1 Participants 54 2.3.2 Equipment 55 2.3.2.1 Video-microscopy Configuration 55 2.3.2.2 Digital Camera Configuration 56 2.3.3 Experimental Procedure 56 2.3.3.1 Staining Procedure 57 2.3.3.2 Digital Camera Procedure 59 2.3.3.3 Video-microscopy Procedure 60 2.3.4 Image Capture Analysis 60 2.3.4.1 Digital Camera Image Capture 61 2.3.4.2 Video-microscopy Image Capture 62 2.3.5 Fungiform Papilla Identification 63 2.3.6 Statistical Analysis 64 2.4 Results 65 2.4.1 Fungiform Papilla Counts with the Digital Camera and Video- microscopy Methods 65 2.4.2 Comparison of Papillae Counts Obtained Using the Digital Camera and Video-microscopy Methods 65 2.4.3 Individual Subject Differences in Fungiform Papillae Counts Between the Two Methods 67 2.4.4 Summary of Findings 70 2.5 Discussion 71

Chapter 3: Quantification of Fungiform Papillae Density During Human Development 74

3.1 General Introduction 75

II 3.2 Aims 76 3.3 General Methods 77 3.3.1 Participants 77 3.3.2 Equipment and Experimental Procedure 77 3.3.2.1 Staining Procedure 78 Part I: Fungiform Papillae Density and Their Distribution Patterns in Children and Adults 80 3.4 Introduction 81 3.5 Aims 81 3.6 Methods 82 3.6.1 Image Analysis 82 3.6.2 Statistical Analysis 84 3.7 Results 85 3.7.1 Overall Variation of Fungiform Papillae Density Between Age Groups 85 3.7.2 Variation of Fungiform Papillae Density Per Cm Intervals Between Age Groups 86 3.7.2.1 Variation of Fungiform Papillae Density in the First Cm of the Anterior Tongue 86 3.7.2.2 Variation of Fungiform Papillae in the Second Cm of the Anterior Tongue 87 3.7.3 Distribution Patterns of Fungiform Papillae in Children and Adults 89 3.7.4 Summary of Findings 95 Part II: Predicting Fungiform Papillae Density 96 3.8 Introduction 97 3.9 Aims 97 3.10 Methods 97 3.10.1 Image Analysis 98 3.10.2 Statistical Analysis 99 3.11 Results 99 3.11.1 Predictor(s) of the Total Fungiform Papillae Count for All Age Groups 99 3.11.2 Summary of Findings 102 Part III: A Comparative Study of Fungiform Papillae Counts of the Anterior Tongue 103 3.12 Introduction 104 3.13 Aims 104 3.14 Methods 105 3.14.1 Image Analysis 105 3.14.2 Statistical Analysis 106 3.15 Results 107

III 3.15.1 Fungiform Papillae Density in Areas X and Y 107 3.15.2 Overall Analysis of Fungiform Papillae Density in Areas X and Y of All Age Groups 107 3.15.3 Summary of Findings 108 3.16 Discussion 109

Chapter 4: Development of a Non-Staining Method for Counting Fungiform Papillae 115

4.1 Introduction 116 4.2 Aims 117 4.3 Methods 117 4.3.1 Participants 117 4.3.2 Equipment 117 4.3.3 Experimental Procedure 118 4.3.3.1 Stage 1; Tongue Wet and No-stain 118 4.3.3.2 Stage 2; Tongue Dried and No-stain 119 4.3.3.3 Stage 3; Tongue Dried, Stained and Dried 119 4.3.4 Image Analysis 121 4.3.4.1 Stage 1 and 2 Analyses 121 4.3.4.2 Stage 3 Analysis 121 4.3.5 Statistical Analysis 122 4.4 Results 123 4.4.1 Description of Fungiform Papillae in Each of the Three Stages 123 4.4.2 Counts of Fungiform Papillae in the Common 0.28 cm2 Circle Area of the Three Stages 127 4.4.3 Summary of Findings 129 4.5 Discussion 130

Chapter 5: The Effect of Chronic Renal Failure on Fungiform Papillae Density in Children 132

5.1 Introduction 133 5.1.1 Renal Failure 133 5.1.2 Effect of Renal Failure on Taste Sensitivity 135 5.1.3 Effects of Renal Failure on Taste Bud Density 136 5.2 Aims 137 5.3 Methods 138 5.3.1 Participants 138 5.3.1.1 Classification of the Patients 138 5.3.2 Experimental Procedure 140 5.3.3 Statistical Analysis 143 5.4 Results 143 5.4.1 Statistical Analysis of Fungiform Papillae Density and eGFR in Control and Patient Groups 143

IV 5.4.2 Topographical Variation of Fungiform Papillae in Control and Patient Groups 145 5.4.3 Summary of Findings 150 5.5 Discussion 151

Chapter 6: General Discussion 153

References 161 Appendix A 189 Appendix B 191 Appendix C 203 Appendix D 205 Appendix E 207

V

I hereby declare that this submission is my own work and, to the best of my knowledge, does not represent the view or work of others, except where duly acknowledge in the text. No part of this thesis has been submitted for a higher degree at another university or institution. This project was approved by the Human Research Ethics Committee, University of Western Sydney, (HEC02/188).

______

Maryam Correa (nee Shahbake)

VI

Acknowledgements

I would like to thank my supervisors, Dr Ian Hutchinson, Professor David Laing, and Dr. Tony Jinks for their support and encouragement. I am forever grateful for all their guidance throughout my studies and without them these studies would not have been possible. Thank you to Carolina Segovia, Fiona Wilkes and Jessica Armstrong for their assistance and motivation.

Many thanks to my parents who have supported and encouraged me throughout my life. A special thanks to my wonderful husband Alejandro, for all his inspirations throughout my studies. Thank you for being the first volunteer, technical wizard and solving everything by clicking two buttons! Thank you to my amazing daughters Jasmine and Jacinta who have showed me anything is possible.

Last but not least, thanks to all the subjects who participated in this study, specially, all the children from Sydney Children’s Hospital, without the participants this study would not have been possible.

VII

List of Figures

Figure 1.1: The human tongue showing the regional location of different types of papillae (Goldstein 1989) 4

Figure 1.2: Drawings of regions containing different types of papillae (Miller 1995) 5

Figure 1.3: (A) Topographical view of a circumvallate papilla showing the papilla surrounded by a moat. (B) Sectional view of a circumvallate papilla showing the location of the taste buds (Gartner and Hiatt 2000) 7

Figure 1.4: Dorsal view of foliate papillae. Arrows indicating the position of taste buds (Gartner and Hiatt 2000) 8

Figure 1.5: (A) Topographical view of filiform papillae surrounding fungiform papillae. (B) Sectional view of filiform papilla (Gartner and Hiatt 2000) 9

Figure 1.6: Drawing of different gustatory papillae. Red arrows show the location of taste buds (Ballard 1997) 10

Figure 1.7: Basic structure of a taste bud (Mistretta 1991; Gartner and Hiatt 2000) 11

Figure 1.8: Anatomy of peripheral taste pathways, including the cranial nerves VII, IX and X (Finger and Silver 1991) 13

Figure 1.9: Embryonic rat tongue and fungiform papilla development. (A) Scanning electron micrographs illustrate the development of fungiform papillae from day 13 to 16 of gestation, Fungiform papilla swelling appears at day 14 and by day 16 the developed papillae are observed. (B) and (C) Scanning electron micrographs and diagrams to illustrate the progressive differentiation in the fungiform papillae development from day 13 to 16 of gestation (Mistretta and Liu 2006) 16

VIII

Figure 1.10: Serial section of fungiform papilla, using H&E staining to show a taste pore using light microscopy (Gartner and Hiatt 2000) 24

Figure 1.11: (A) and (B) show the 9 mm2 area investigated by Segovia et al., (2002) and Shahbake (2002), respectively. (C) Shows the 6 mm diameter area analysed by Shahbake (2002) 27

Figure 1.12: (A) The video-microscopy configuration. (B) Shows the position of the TV and subject in the amended video-microscopy procedure which reduced tongue movement in children (Shahbake 2002) 28

Figure 1.13: (A) Contact endoscopy configuration. (B) The contact endoscope on a stained tongue surface (Just et al., 2006) 30

Figure 1.14: (A) Image of smooth egg-shaped fungiform papillae of a healthy participant with the use of a video microscope and (B) showing abundant straight blood vessels running parallel to each other in a fungiform papilla using contact endoscopy. (C) Image of flat and irregular fungiform papillae of a participant (video microscope) suffering chorda tympani injury and (D) shows the atrophic, irregular and less prevalent blood vessels (contact endoscopy) (Negoro et al., 2004) 31

Figure 1.15: The classification of fungiform papillae and blood vessels (Negoro et al., 2004) 32

Figure 2.1: The video-microscopy configuration (Shahbake 2002) 55

Figure 2.2: The Nikon Coolpix 4500 camera with the macro light and rotating head 56

Figure 2.3: Diagram of the staining procedure with blue food dye 58

IX

Figure 2.4: (A) Demonstrating the seating position of the participant. (B) The tongue showing the stained area and the 1 cm filter paper scale on the other side of the tongue 59

Figure 2.5: Image of the tongue captured by the digital camera. Arrows show fungiform papillae in the stained area at X21 magnification 61

Figure 2.6: Image of the tongue captured by the video-microscopy procedure. Arrows show fungiform papillae in the stained area at X30 magnification 62

Figure 2.7: Examples of different shaped fungiform papillae using the video- microscopy method: (A) “double papilla”, (B) flat papillae, (C) short and wide necked mushroom shaped papilla (Shahbake 2002). The digital camera method: (D) “double papilla” and flat papillae, (E) raised and conical shaped papilla, (F) short and wide necked papilla 64

Figure 2.8: Interaction plot showing interactions between the two factors (Method versus Age) 66

Figure 2.9: Fungiform papillae counts for individuals using the digital camera and the video-microscope for (A) Children and (B) Adults. Arrows show large differences in counts between the two methods for four participants 68

Figure 2.10: Comparison of images from the two photographic methods of the same area showing the visibility of the fungiform papillae of participant 4 69

Figure 3.1: Stepwise diagram of the staining procedure 79

Figure 3.2: Diagram demonstrating preparation of the images before counting fungiform papillae 83

X

Figure 3.3: Interaction plot showing variation between males and females in different age groups 87

Figure 3.4: The Tongue of (A) 7-8 year old, (B) 9-10 year old, (C) 11-12 year old and (D) an adult. Arrows show fungiform papillae 88

Figure 3.5: Examples of different distribution patterns showing drawing of tongues of two 8 year olds. The dotted line indicates the end of the second cm. (A) Type A, arrow showing cluster of fungiform papillae (B) Type B, evenly distributed fungiform papillae 89

Figure 3.6: Diagram of different distribution patterns in the adult group, the dotted line is an indication of the end of the second cm. (A) Type A, arrow showing cluster of fungiform papillae. (B) Type B, evenly distributed fungiform papillae 90

Figure 3.7: Tongues of 7-8 year old children showing different distribution patterns. (A) Type A, arrow indicating cluster of fungiform papillae (FP) and (B) Type B, arrow showing no cluster of FP; scale= 1 cm 91

Figure 3.7 (Cont.): Tongues of 9-10 year old children showing different distribution patterns. (C) Type A, arrow indicating cluster of fungiform papillae (FP) and (D) Type B, arrow showing no cluster of FP; scale= 1 cm 92

Figure 3.7 (Cont.): Tongues of 11-12 year old children showing different distribution patterns. (E) Type A, arrow indicating cluster of fungiform papillae (FP) and (F) Type B, arrow showing no cluster of FP; scale= 1 cm 93

Figure 3.7 (Cont.): Tongues of adults showing different distribution patterns. (G) Type A, arrow indicating cluster of fungiform papillae (FP) and (H) Type B, arrow showing no cluster of FP; scale= 1 cm 94

Figure 3.8: (A) The image of the stained area was sub-divided into small areas. (B) A diagram of the tongue showing the numbering sequence of the 8

XI

small areas 98

Figure 3.9: Fungiform papillae counts in each of the 8 small areas in each of the groups 101

Figure 3.10: Images of the tongue showing the 2 x 1 cm bands (Part I), 8 small areas (Part II) and two 0.28 cm2 circles superimposed on an image showing Areas X and Y (Part III) 106

Figure 4.1: The Canon AF SLR (EOS-1Ds) camera with attachments 118

Figure 4.2: Diagrams of the procedures in the three Stages of tongue preparation for photography 120

Figure 4.3: The image preparation before the fungiform papillae count in (A) Stages 1 and 2 (B) Stage 3 122

Figure 4.4: Fungiform papillae of participant in (A) Stage 1, (B) Stage 2 and (C) Stage 3. In (D) and (E) arrows show hidden fungiform papillae in Stage 2. (F) Dark stained papillae and surrounding tissue and (G) light stained papillae and surrounding tissue in Stage 3 124

Figure 4.5: Types of blood vessel patterns of fungiform papillae. (A) Dotted pattern and (B) branched pattern. Arrows indicate blood vessels 125

Figure 4.5 (cont.): (C) Branched pattern and (D) branched and lopped pattern. Arrows indicate blood vessels 126

Figure 4.6: B-A difference plot of fungiform papillae counts between Stage 1 and 2 128

Figure 4.7: B-A difference plot of fungiform papillae counts between Stage 1 and 3 128

XII

Figure 4.8: B-A difference plot of fungiform papillae counts between Stage 2 and 3 129

Figure 5.1: Diagram of hemodialysis set up (Martini 2006) 134

Figure 5.2: Diagram of peritoneal dialysis set up (Martini 2006) 135

Figure 5.3: Diagram of the procedure for superimposing the 0.28 cm2 circle on the tongue using the 1 cm scale 142

Figure 5.4: Relationship between eGFR and fungiform papillae density of patient and control groups 144

Figure 5.5: Tongue of patient 10. White arrow shows yellow fur coating on tongue surface 146

Figure 5.6: Typical tongues of (A) patient and (B) control. The white dotted circle indicates the area within which fungiform papillae density was measured. Arrows show fungiform papillae. Scale = 5 mm 147

Figure 5.7: Tongue of patient 12. White arrows show examples of enlarged and irregular fungiform papillae 149

XIII

List of Tables

Table1.1: Summary of fungiform papillae density obtained using different procedures 34 Table1.2: Summary of taste pore density obtained using different procedures 38 Table 3.1: Counts of total fungiform papillae and per cm intervals of all groups 85 Table 3.2: The predictor models of total fungiform papillae of the anterior tongue of 7-8 year olds 100 Table 3.3: The predictor models of total fungiform papillae of the anterior tongue of 9-10 year olds 100 Table 3.4: The predictor models of total fungiform papillae of the anterior tongue of 11-12 year olds 100 Table 3.5: The predictor models of total fungiform papillae of the anterior tongue of adults 101 Table 3.6: The overall predictor models of total fungiform papillae of the anterior tongue of all age groups 101 Table 5.1: Description of patient and clinical control groups 139 Table 5.2: Clinical characteristics of age and gender matched patients and controls 140 Table 5.3: Tongue topographical characteristics of age and gender matched patients and controls 148

XIV

List of Abbreviations

ANOVA Analysis of Variance AT Anterior Tongue B-A plot Bland–Altman plot BDNF Brain Derived Neurotrophic Factor BMTN Bone Marrow Transplant Nephropathy BRF British Renal Foundation CARI Caring Australia with Renal Impairment guidelines CC Cell Counter CE Contact Endoscopy ˚C Centigrade degree cm Centimetre CRF Chronic Renal Failure CT Cadaver Tongue eGFR estimated Glomerular Filtration Rate EM Electron Microscopy F Female FP Fungiform Papillae FSGS Focal Segmental Glomerulosclerosis H&E Hematoxylin and Eosin HSP Henoch-Schonlein Purpura K/DQOI Kidney Disease Outcome Quality Intuitive guideline LM Light Microscopy LoA Limits of Agreement M Male MG Magnifying Glass μm Micrometer μs Microsecond Min Minute mL Milliliter mm Millimetre

XV

MT Medium-taster NaCl Sodium Chloride NGF Nerve Growth Factor NIH National Institute of Health NT Non-taster NT3 Neurotrophin-3 NT4 Neurotrophin-4 P Pink Pap Papilla PP Pale Pink PROP 6-n-propylthiouracil PTC Phenylthiocarbamide PUV Posterior Urethral Valves RN Reflux Nephropathy RPGN Rapidly Progressive Glomerulonephritis S.D. Standard Deviation S.E. Standard Error SEM Scanning Electron Microscopy ST super-taster Tb Taste Bud TP Taste Pore TrkB Tyrosine Kinase B UTI Urinary Tract Infection VM Video Microscopy Zn Zinc

XVI Abstract

Abstract

The specific objectives of this thesis were (1) develop a rapid and portable method to quantify fungiform papillae on the anterior tongue of children and adults, (2) determine how fungiform papillae density changes through childhood to adulthood, (3) determine whether a particular region of the tongue is a reliable predictor of the fungiform papillae density of the anterior tongue, (4) validate the earlier work of others in quantification of fungiform papillae in small regions of the anterior tongue, (5) develop a non-invasive method for counting papillae, and (6) determine whether chronic renal failure (CRF) affects fungiform papillae density.

The first aim was achieved by using a digital camera, which was validated with the previous non-invasive and non-portable method of video-microscopy (Miller and Reddy 1990 a, b). No significant difference in the results were obtained using both methods (Shahbake et al., 2005) (Chapter 2). The new method was also less time consuming in image capturing in comparison to video-microscopy and most importantly it was portable. Indeed the portability of this method made it possible to conduct the studies at schools and different locations at the University and at Hospital Outpatient Clinics ensuring an adequate number of participants were recruited. In addition, it provides a suitable procedure that can be utilized concurrently with psychophysical studies that involve large numbers of participants and requires short times for data collection.

According to the literature there are many differences between adult and children in their taste sensitivity, perceptions and anatomy (Miller and Reedy 1990b; Stein et al., 1994; Segovia et al., 2002). However no study has reported at what age these differences disappear from childhood to adulthood. Thus the second aim of this study was to investigate the variation in fungiform papillae density from childhood to adulthood and determine at what age any differences disappear (Chapter 3: Parts I to III). The flexibility

XVII Abstract and portability of the digital camera method was demonstrated in Chapter 3 which involved a large number of participants (n= 115) from children as young as 7 years old to adults. The participants were divided into four groups, (1) 7-8 year olds, (2) 9-10 year olds, (3) 11-12 year olds and (4) adults. Fungiform papillae were counted in the anterior tongue (tip to 2 cm towards posterior). The results indicated that, fungiform papillae densities were highest in the 7-8 year olds, gradually decreasing until 9-10 years of age where adult levels were reached. This is the first study that has demonstrated that fungiform papillae density, reaches adult numbers by mid- childhood (Correa et al., 2013). There were also variations in the fungiform papillae distribution patterns between children and adults. The fungiform papillae in children were mainly arranged in clusters whereas in adults these were more evenly distributed.

The findings of Chapter 3, Part I were further studied in Part II where the aim was to determine whether a single small area of the anterior tongue could predict fungiform papillae density on the whole of the anterior tongue. Overall the best predictor area in children aged 7-10 year olds was at the tip of the tongue (Area 1), whereas in 11-12 year olds and adults the mid-region of tongue (Area 5) was the best predictor. Thus by using a digital camera to rapidly obtain images of papillae and a single small area of the tongue, a highly efficient procedure was achieved. According to the results, fungiform papillae maturity is attained in children aged 9-10 years (Correa et al., 2013). The captured images from Chapter 3, Part I were also used in Part III, to check the reliability of previously arbitrary locations by others, which near the tip of the tongue were chosen, as a reliable location for fungiform papillae quantification. The results indicated there were strong correlations between the total fungiform papillae density on the anterior tongue from childhood to adulthood and the area chosen, thus, verifying previous studies (Correa et al., 2013).

Chapter 4 was undertaken in an effort to overcome the discomfort of patients in clinical settings during the staining and drying procedures used before the images of the tongue are captured. In this experiment the tongue images of

XVIII Abstract healthy volunteers were captured in three stages (wet, dried and stained) and fungiform papillae density counted independently at each stage. The results showed that there were no significant differences between the fungiform papillae counts in the three stages. Importantly, the papillae were easier to identify when the tongue was wet in its natural state, rather than dried, because they were more elevated. Furthermore, various types of blood vessels were observed on the surface of the papillae, in accordance with a previous study by Negoro et al., (2004). Overall, it was concluded that it is possible to identify and count fungiform papillae using a high-resolution digital camera without staining the tongue surface or treating the tongue in any way to obtain images (Correa et al., 2015).

Subsequently, in Chapter 5 the non-staining procedure and the high- resolution digital camera method were used to investigate the effect of CRF on the fungiform papillae density of children aged 5-17 years. The main finding indicated a correlation between fungiform papillae density and kidney function. In addition, the papillae density was significantly lower in patients than their matched controls (Correa et al., 2015). Importantly, it was concluded that the loss of fungiform papillae might be one of the first signs of taste degeneration in CRF patients.

In conclusion the study demonstrated that there were differences in fungiform papillae characteristics in children and adults and that a digital camera provides a convenient tool for achieving these findings. Children aged 11-12 years had similar fungiform papillae densities and predictor areas as adults and it was concluded that fungiform papillae maturity was reached at this age. The results also identified the best predictor of fungiform papillae density for children and adults. Additionally, the methods used by other earlier studies were validated, and the results demonstrated that small tip regions of the anterior tongue were suitable indicators of fungiform papillae density. Importantly, a non-staining method was developed to count papillae and the new method was used in a clinical setting, where it was shown that, patients with CRF had significantly lower fungiform papillae densities than their clinical controls. Overall, the results indicate that a digital camera

XIX Abstract provides a simple and convenient means for conducting anatomical studies on the anterior tongue of humans both in the laboratory and clinical settings.

XX Chapter 1

Chapter 1

General Introduction

1 Chapter 1

Introduction

Senses are the physiological methods of perception and are systems that consist of a group of cell types that respond to a specific kind of physical energy. They stimulate a defined region within the brain where the signals are received and interpreted. In contrast to vision or hearing the senses of gustation (taste) and olfaction (smell) are called chemical senses and depend on chemical stimuli that are present in food and drink or in the air. The sense of taste protects us from unsafe foods and helps to maintain a consistent chemical balance in the body (Schmidt and Thews 1980). For example, liking sugar or salt satisfies the body’s need for carbohydrates and minerals. The sensory receptors for taste are within taste buds, which are primarily located in the oral cavity. Every taste bud contains cells that detect five primary tastes (sour, sweet, bitter, salty and umami) (Lindemann 1996; Koehler and Leonhaeuser 2008).

Human taste preferences change throughout a person’s life. A food that might seem unpleasant and unacceptable in childhood can be incredibly enjoyable and acceptable later in life as an adult. For example, adults enjoy the bitterness of caffeine in the daily cup of coffee, while it is an unpleasant flavour to children. According to the current literature there are many differences in human food preferences, perception and behaviour from birth to adulthood and into old age. In addition, anatomical and physiological differences occur during childhood through to adulthood. Currently there is very limited information about the development of the human gustatory system during different stages of childhood. Thus, one of the major aims of this research was to examine a large group of participants from childhood to adulthood and discover at what ages anatomical differences exist in the peripheral taste system. For this research to be possible a portable, fast and cost effective method was required. In reviewing the literature there was no such method available to quantify the number of taste papillae, which contain taste buds. Therefore, the first aim of this research was to use today’s digital technology to develop such a method.

2 Chapter 1

In addition to the aging process, there are number of other aspects that also effect the taste system, for example, illness and medical intervention. Although there are many reports in the literature (Zheng et al., 2002; Sandow et al., 2006) that suggest that diseases such as diabetes, CRF and cancer affect human taste preferences and perception, only very limited reports indicate what happens to the anatomy and physiology of the human peripheral taste system (i.e. taste papillae). Therefore, another aim of this research was to investigate the effects of a common disease, CRF, on the human peripheral taste system. The following is a brief review of what is known about the anatomy and development of the gustatory system, the methods developed to explore the peripheral gustatory system and the effect of diseases on the system.

3 Chapter 1

1.1 Anatomy of the Peripheral Gustatory System

Since one of the main objectives of this thesis is to study the developmental changes in fungiform papillae on the anterior tongue, below is a review of the general anatomy of the gustatory system, with particular reference to different taste papillae structures and their functions.

1.1.1 The Tongue

The tongue consists almost entirely of muscle fibres. The movement of the tongue against the roof of the mouth, teeth and lips help to shape vocal sounds into words, as well as to eat and drink effectively (Martini 2006; Peck and Hannam 2007). The base of the tongue is located in the throat just above the larynx (voice box) and extends to the hyoid bones (Peck and Hannam 2007). The tongue is wrapped in the lingual membrane and is covered with small structures called papillae (Figure1.1), which contain taste buds and taste receptor cells (Just et al., 2005). These papillae are responsible for the tongue’s textured surface (Martini 2006).

Figure 1.1: The human tongue showing the regional location of different types of papillae (Goldstein 1989)

4 Chapter 1

1.1.2 Taste Papillae

Marcello Malpighi first discovered taste papilla in 1664 (Bradley and Stern 1967). A taste papilla can be seen as a small red dot or a raised bump on the surface of the tongue (Just et al., 2006). The papillae are categorised into three different types of gustatory papilla (i.e. contain taste buds) (Figure 1.2), circumvallate, foliate and fungiform papilla, and two non-gustatory papillae (no taste buds), the conical and filiform papilla (Kullaa-Mikkonen and Sovari 1985; Just et al., 2006).

Figure 1.2: Drawings of regions containing different types of papillae (Miller 1995)

1.1.2.1 Fungiform Papilla

Fungiform papillae are mushroom-like structures (Figure 1.2), which vary in appearance and are distributed over a large area of the tongue (Miller 1995). Human fungiform papillae appear as red spots on the tongue because they are richly supplied with blood vessels (Negoro, et.al, 2004). Most papillae on the tip of the tongue are shorter than those on more posterior regions and become taller and larger towards the back of the tongue. Some papillae on the margin of the tip are elongated like conical papillae and generally lack

5 Chapter 1 taste buds. Interestingly Mistretta and Baum (1984) reported 12.8% of fungiform papillae do not contain taste buds, whilst Miller and Reedy (1990a), reported a range of 3 to 35%. Miller and Reedy (1990a, b) studied the distribution of human fungiform papillae and reported an overall average density of 41.1 and 43.8 fungiform papillae/ cm2, respectively, on the tip of the tongue. The latter results were supported by Shahbake et al., (2005) who found 42.4 and 45.7 papillae/ cm2 using two different counting procedures (Chapter 2). However, a more recent study by Just et al., (2006) reported lower fungiform densities of 28.3 and 15.6 papillae/ cm2 on the edge of the tip of the tongue. The large age range of the subjects (7 to 71 years) may have been the reason for the low densities. For example, it has been reported that children have higher fungiform papillae density than adults (Mochizuki 1939; Segovia et al., 2002; Shahbake 2002).

The number of taste buds is reported to differ among fungiform papillae and there are large variations among subjects in the distributions of fungiform papillae and taste buds (Miller 1995; Miller and Bartoshuk 1991). For example, children have higher papillae densities on the tip of the tongue than adults (Segovia et al., 2002; Shahbake 2002). Thus, Segovia et al., (2002) reported 91 papillae/ cm2 in 8 to 9 year old males and 68 papillae/ cm2 in adults (18 to 35 year old), which was supported by Shahbake (2002), who reported 104 papillae/ cm2 in 8 to 9 year old males and females.

6 Chapter 1

1.1.2.2 Circumvallate Papillae

Circumvallate (vallate) papillae are the largest papillae located on the surface of the tongue. They appear sunken, flattened and have an oval shape with a trough separating them from the surrounding wall (Figure 1.3A) (Emura et al., 2001) and are found at the back of the tongue in a V-shaped line across the root of the human tongue (Miller 1995). The taste buds are in tiers within the trough of the papilla and are situated on the circumvallate line (Figure 1.3B).

TASTE BUD

(A) (B) Figure1.3: (A) Topographical view of a circumvallate papilla showing the papilla surrounded by a moat. (B) Sectional view of a circumvallate papilla showing the location of the taste buds (Gartner and Hiatt 2000)

1.1.2.3 Foliate Papilla

Foliate papillae appear as a series of folds on the posterior lateral sides of the tongue and contain taste buds (Figure 1.4). The word, foliate means “leaf like” which is an indication of their appearance. The folds are separated by grooves of different depths perpendicular to the longitudinal axis of the tongue in an extent of 1 to 2 cm (Sicher 1966). The number of human foliate papillae was first reported by Hou-Jensen (1933) and confirmed by Mochizuki (1939). Both investigators reported a range of 2 to 9 foliate papillae on each side of the tongue. The shape and arrangement of human foliate papillae differs among individuals. Frequently an asymmetric arrangement is visible.

7 Chapter 1

For example, on one side of the tongue the papillae may be developed differently from those on the other side. Occasionally they are developed on one side only, and sometimes in large numbers and sizes and this abnormality is sometimes mistaken for a tumour (Svejda, and Janota 1972). The foliate papillae are morphologically and functionally similar to the circumvallate papillae, but they undergo changes with age especially in humans (Svejda, and Janota 1974).

TASTE BUD

Figure 1.4: Dorsal view of foliate papillae. Arrows indicate the position of taste buds (Gartner and Hiatt 2000)

1.1.2.4 Conical Papilla

Conical papillae are non-gustatory papillae and do not contain taste buds. The papillae are cylindrical at the base and end in a sharp point in human (Miller and Bartoshuk 1991). Kutuzov and Sicher (1951) conducted a histological examination of conical and filiform papillae on the tongue of 21 albino rats aged 10 days to 5 months and categorised conical papillae into two groups, namely, simple and giant conical papillae. The anterior tongue contained simple conical papillae and the connective tissue core was in a shape of a condensed cone. Conical papillae become larger in size and density toward the back of the tongue and were categorised as giant conical papillae (Kutuzov and Sicher 1951). The epithelial structure of the giant conical papillae does not differ from the simple conical papilla in rats

8 Chapter 1

(Kutuzov and Sicher 1951) and their main function is prehension. For example, grasping food particles on the surface of the tongue of primates and insectivores (Matsukawa and Okada 1994).

1.1.2.5 Filiform Papillae

Filiform papillae are also categorised as non-gustatory papillae because they lack taste buds (Goldstein 1989; Miller and Bartoshuk 1991). This category of papilla has a finger-like shape and bent appearance (Just et al., 2006) with a tail (fila), which extends from the apex in human. Filiform papillae have an abrasive function during mastication (Goldstein 1989; Matsukawa and Okada 1994) and are found over the entire surface of the tongue giving its rough appearance in human and primates (Figure 1.5).

FILLIFORM FILLIFORM

* FUNGIFORM

(A) (B) Figure 1.5: (A) Topographical view of filiform papillae surrounding fungiform papillae. (B) Sectional view of filiform papilla (Gartner and Hiatt 2000), *CT= connective tissue

1.1.3 Taste Buds

Taste buds are clusters of specialized epithelial cells that primarily reside in three distinct locations within the lingual epithelium; namely the circumvallate, foliate and fungiform papillae (Figure 1.6) (Witt and Kasper 1999), in the palate, back of the mouth, pharynx, epiglottis and larynx (Plattig 1984; Nelson 1998). Human taste bud diameters range from 25 to 60 and 40 to 70 m (Bradley 1972; Plattig 1984; Miller and Reedy 1990a), and consist of approximately 50 to 150 taste cells, that include supporting (sustentacular)

9 Chapter 1 cells, basal cells and receptor cells (Figure 1.7). The taste cells are subdivided into four types; type I (dark), II (light), III (intermediate) and IV (basal). The basal cells are at the base of the taste bud; they divide to produce post-mitotic light, intermediate and dark taste cells. Dark cells are defined by a dark cytoplasm and dense-core granules (small vesicle with a dark center) at the tip of the cell providing the mucous material that fills the taste pores. Light cells are characterized by a light cytoplasm, clear vesicles and mitochondria in the tip of the cell and the intermediate cells have characteristics that are intermediate between light and dark cells. The intermediate cells have synapses to nerve fibers, therefore, are considered to be receptor cells (Murray and Murray 1971; Nelson 1998). Delay et al., (1986) reported that these cell types represent a progression through different developmental stages from dark (type I) to intermediate (III) to light (II), but other investigators suggest independent cell cycles for the different types (Farbman 1980).

Figure 1.6: Drawing of different gustatory papillae. Red arrows show the location of taste buds (Ballard 1997)

Each taste bud is oval and opens to the epithelial surface via a small opening called a taste pore. The taste pore is a small channel of about 2 to 7 m diameter (Arvidson 1976; Plattig 1984; Just et. al, 2005). From the pore protrude the microvilli (taste hairs), which arise from the tips of individual taste cells (Figure 1.7) and are 2 to 3 m long with a diameter of 0.1 to 0.2 m. Microvilli extend from the surface of the taste cell and are exposed to tastants in the saliva on the tongue. It is here, that sensory transduction takes place (Guyton 1987; Nelson 1998). Thus, the presence of taste pores signifies that receptor cells are present and functioning.

10 Chapter 1

Figure 1.7: Basic structure of a taste bud (Mistretta 1991; Gartner and Hiatt 2000)

1.1.3.1 Taste Bud Distribution

The pattern of taste bud distribution over the tongue surface is similar among humans and other mammals (Bradley 1971). Taste buds also exist in regions of the oral cavity other than the tongue, and are known as extra-lingual taste buds (Miller 1995), which are reported in soft palate (Nilsson 1979), larynx (Wilson 1905), epiglottis, laryngeal pharynx and oral pharynx (Lalonde and Eglitis 1961). Also there have been reports of taste buds in the human adult oesophagus (Burkl 1954), near the openings of sublingual salivary ducts in some primates (Hofer 1977) and near the ducts of molar glands in rodents (Iida et al., 1983). Miller and Smith (1984) estimated that about 25% of the total taste buds of the hamster are extra-lingual and a similar proportion was reported by Mistretta (1991) in rats. Although there have been a number of studies on extra-lingual taste buds, it is not known whether their function differs from those on the tongue.

The number of taste buds per papilla varies between different types of papillae. According to Hou-Jensen (1933) and Mochizuki (1939), an average of 240 taste buds occur in a circumvallate papilla and 118 taste buds in a

11 Chapter 1 foliate papilla, whereas, fungiform papilla only contain 0 to 22 taste buds (Miller and Reedy 1990a; Segovia et al., 2002). Variations in taste bud densities also occur in different locations of the tongue. For example, Miller (1986) reported that in humans taste bud density is highest at the tip of the tongue (116 taste bud/ cm2) in comparison to the mid region of the tongue (25.2 taste bud/ cm2).

1.1.3.2 Taste Bud Innervation

The sensory innervation of the tongue plays a major role in taste perception (Ballard 1997). It provides information in regards to different taste qualities (i.e. sweet, salt, sour, bitter and umami), temperature (cold and warm) and consistency (hard and soft) of food or beverages (Finger and Silver 1991). The sensory nerve supply to the human tongue is provided by three different cranial nerves, namely, the facial (VII), glossopharyngeal (IX) and vagus (X) nerves. Each cranial nerve is subdivided into branches that relate to different parts of the body (Brand 1989). For example, the branches of the facial nerve that relates to the gustatory system are the chorda tympani and greater superficial petrosal branch.

Each papilla type has a distinctive innervation. Taste buds of fungiform papillae are innervated by a branch of the facial nerve called the chorda tympani, whilst the circumvallate papillae and their taste buds are innervated by the lingual branch of the glossopharyngeal nerve. Foliate papillae and their taste buds are innervated by both the chorda tympani (anterior papilla folds) and glossopharyngeal (posterior folds) nerves (Mbiene et al., 1997) (Figure 1.8).

12 Chapter 1

Figure 1.8: Anatomy of peripheral taste pathways, including the cranial nerves VII, IX and X (Finger and Silver 1991)

13 Chapter 1

1.2 Development of the Peripheral Gustatory System

Most studies of taste buds and taste papillae have been undertaken on fungiform and circumvallate papillae. The development of these structures is described below for the various stages of life.

1.2.1 Foetal and Neonatal Development

In reviewing the literature there are many animal models that have been used for developmental studies of the mammalian taste system. This has occurred even though the gestation period of some animals (e.g. rats, 21 days) are very short in comparison to that of humans (9 months). Despite this difference, useful comparisons have been made between species.

1.2.1.1 Animal Studies

Fungiform Papillae

Mistretta (1972) studied the development of the rat tongue and palate with the techniques of scanning electron microscopy and light microscopy. Observations were made from foetuses of gestational age 14 to 21 days and from neonates aged 1 to 20 days. It was reported that fungiform papillae and circumvallate papillae appeared early in the stages of tongue development and before the palates are defined. Fungiform papillae first appear on the foetal rat tongue between 14 and 15 days of gestation (Mistretta 1991; Mbiene et al., 1997), and consist of clusters of cells of the epithelium (Figure 1.9). The papillae appear similar to developed human fungiform papillae (Mistretta 1991). By 20 days gestation the morphology of the fungiform papillae differs from earlier days and consists of a core of connective tissue covered by lingual epithelium. The latter is the first stage that taste buds are being formed, although they do not mature until 12 days after birth (Mistretta

14 Chapter 1

1972; Hill 2001; Mistretta and Hill 2003; Mistretta and Liu 2006). The mature taste buds contain cells which are oriented longitudinally and stain distinctly from the surrounding epithelium. One of the final morphological events in the formation of a taste bud is the appearance of the pore (Mistretta 1972; Mbiene et al., 1997; Sharaby et al., 2006). Developing taste buds appear to be sealed off until a taste pore forms to expose the apical tips of the taste receptor cells to chemical stimulants in the oral environment. Thus Oakley et al., (1991) defined a mature taste bud as one that has developed a taste pore. A mean of 10.5 days is required for the maturation of new taste buds in rats (Oakley et al., 1991), which is similar to the development time for hamster circumvallate taste buds (Belecky and Smith 1990).

However, taste buds do not mature simultaneously (Mistretta 1972). When 90 fungiform papillae from the tongues of one, four, eight, ten, and twelve day old neonatal rats were examined, only a single taste pore was found on the tongue of the one-day-old rat. Six pores were counted in the 4 day old rat, 32 in the 8 day old and by 12 days of age there were 67 pores in 93 papillae. It was not until 14 days after birth that each fungiform papilla contained a taste pore.

Similarly, Bradley and Stern (1967) reported that in the sheep presumptive taste buds first appeared at 50 days of gestation and by 100 days the buds appeared morphologically mature. The development time for the sheep taste bud is similar to that of humans. Initially at 7 weeks in the human foetus, collections of cells resembling taste buds can be observed and they are mature by weeks 13 to 15 (Bradley and Stern 1967). The 7 weeks maturation period of human taste buds is equivalent to the 50 to 100 days of gestation in sheep. Although human and sheep taste buds have mature morphological appearances before birth, their development is not completed until after birth.

15 Chapter 1

(A) (B) (C)

Figure 1.9: Embryonic rat tongue and fungiform papilla development. (A) Scanning electron micrographs illustrate the development of fungiform papillae from day 13 to 16 of gestation, fungiform papilla swelling appears at day 14 and by day 16 the developed papillae are observed. (B) and (C) Scanning electron micrographs and diagrams to illustrate the progressive differentiation in the fungiform papillae development from day 13 to 16 of gestation (Mistretta and Liu 2006)

Circumvallate Papillae

Interestingly, the rat tongue has a single circumvallate papilla that lies on the dorsal midline. It consists of a 1 mm diameter island of tissue surrounded by a trench. The several hundred taste buds in the circumvallate trench walls mature during a postnatal development period of 90 days (Farbman 1980; Oakley et al., 1991) and the papillae begin to form on the foetal rat tongue between days 14 to 15 of gestation (Oakley 1988; Oakley et al., 1991) that resembles a mound of cells. The circumvallate papilla is well defined by day 16 although the walls that surround the adult papilla have not been developed (Krimm and Hill 1998). The surrounding wall appears on day 18 of gestation and at one day after birth the papilla has an adult topographical appearance. Taste bud development in the circumvallate papilla is first seen

16 Chapter 1 at 20 days of gestation and by the 12th day after birth, many taste buds are found in the trench walls of the papilla (Mbiene et al., 1997).

1.2.1.2 Human Studies

Bradley (1972) studied the development of gustatory papillae and taste buds in the human foetus and divided the development of the gustatory system into three stages; preformatory, formatory and differential stages. The preformatory stage (week 4) is the initial tongue formation. The tongue is developed to a series of paired and unpaired swellings arising from the branchial arches in the floor of the mouth cavity. In the formatory stage (weeks 5 and 6) the swellings of the preformatory stage change dimensions and begin to fuse, the tongue becomes broader and lower and separates from the lower jaw. Later, in the differential stage (from week 7) the foetal tongue begins to resemble the adult tongue and this is the stage that the gustatory papillae begin to develop.

Subsequently, the development of papillae can be divided into two phases. The first phase is the morphogenesis phase, in which the papillae structures are produced. The second phase is the innervation phase in which the taste buds develop. The innervation phase therefore, represents a maturation or functional phase (Bradley 1972; Oakley et al., 1991).

One of the first comprehensive studies that examined the development of papillae in the human foetus was by Bradley and Stern (1967) who studied 70 specimens of the human foetal tongue with the use of light microscopy. It was reported that at 6 to 7 weeks the tongue epithelium consists of a superficial and a deep layer of cells. Collections of cells resembling taste buds were first seen at week 7 and by week 11 the presumptive bud was formed. Between weeks 12 to 14 the cells of the presumptive bud elongate, pierce the surface as a small tuft of cells and the surrounding surface epithelium forms the taste pores. The mature taste bud is identified at weeks 13 to 15 (Bradley and Stern 1967; Hersch and Ganchrow 1980). Taste hairs

17 Chapter 1 were reported to be present on the apical end of bud cells and extend into the base of the pore (Hersch and Ganchrow 1980). From weeks 16 to 20 some taste buds continue to develop in complexity and some contain multiple taste pores.

Scanning electron microscopy was used by Hersch and Ganchrow (1980) to study the development of gustatory papillae in 15 cadaver human tongues from the first trimester (weeks 8 to 26). The fungiform papillae were first observed from weeks 10 to 13 on the dorsal surface of the anterior tongue, which was covered with papillary structures. Some of the fungiform papillae also contained an opening with an average diameter of 3 to 5 m. During weeks 23 to 26 the papillae were surrounded by distinct and separate filiform papillae and some fungiform papillae contained definite taste pores. The latter results were consistent with previous research by Arvidson (1976).

Later, Witt and Reutter (1997) examined development of 32 human foetal tongues obtained during weeks 6 to 15 of gestation, and followed the development of fungiform papilla and circumvallate papilla using scanning electron microscopy. The first evidence of fungiform papillae was a series of epithelial swellings on the anterior tongue during week 7. By week 10 many fungiform papillae revealed depressions in their apical surfaces, and at 12 to 13 weeks these thin epithelial cells disappeared and microvilli were observed. The first circumvallate papilla appeared in week 6 as a single median papilla and 4 to 5 circumvallate papillae appeared on each side of the sulcus terminalis by weeks 8 to 10. The first pit-like grooves in the apical surface of the circumvallate papilla appeared at weeks 6 to 7. Filiform papillae were not observed during the first trimester, whereas fungiform papillae were widely distributed. Later studies by Witt and Kasper (1999) and Jung et al., (2004) supported the latter findings and reported that pores were observed at postovulatory weeks 14 and 15 in fungiform papillae respectively. According to Arvidson (1979), Witt and Reutter (1997) and Witt and Kasper (1999) foetal human fungiform papillae (up to week 15) contain a maximum of 2 taste buds.

18 Chapter 1

1.2.2 Infant to Adult Development

Many studies have shown that the density of taste buds and fungiform papillae changes from childhood to adulthood (Mochizukie 1939; Schieber 1992; Segovia et al., 2002; Just et al., 2006), however, no studies have reported at what age development ceases. The following are brief reviews of the literature as regards to differences in the density of taste papillae and taste buds from childhood to adulthood.

1.2.2.1 Animal Studies

Light microscopy was used to investigate whether there were any anatomical differences in taste bud density with age in 38 Fischer rats aged 4 to 37 months (Mistretta and Oakley 1986). The animals were divided into 3 groups and they quantified the number of fungiform papillae that contained taste buds and the diameter of the taste buds. The mean percentage of fungiform papillae that contained taste buds in the three groups was 99.6% (4 to 6 months), 99.3% (20 to 24 months) and 94.7% (30 to 37 months). The taste bud diameter ranged from 58 to 59 m, with no significant differences between the groups. In the latter study, the tongues were also divided into four quarters and the number of papillae counted. The papillae distribution was 49% on the anterior quarter of the tongue, 23% on the next quarter and 28% on the posterior two quarters, which was very similar to that of the normal rat population of fungiform papillae reported by Miller and Preslar (1974). The latter authors reported 50% of total papillae were on the anterior quarter, 28% on the next quarter and 22% on the posterior half of the tongue.

Similarly Mistretta and Baum (1984) found no age-related differences in the number of taste buds in fungiform and circumvallate papillae in 24 Wistar rats (12 young and 12 adults) aged 6 to 24 months. Also there were no significant difference between the numbers of taste buds in papillae on the posterior part of the tongue and the anterior part of the tongue. To quantify papilla size each tongue was divided into successive quarters, from the circumvallate

19 Chapter 1 papilla to the tip. The most anterior quarter was divided into two due to the high fungiform papilla density. Other than changes in size no differences in papilla or taste bud morphology were apparent with the light microscope. The fungiform papilla density was reported to be 116 and 113 fungiform papillae per tongue in young and old rats, respectively. The only difference reported was that fungiform papillae were larger in the older rats.

Another study by Bradley et al., (1985) reported an average of 4 to 5 of taste buds per fungiform papilla in 15 Rhesus monkeys aged 4 to 31 years. There were no significant differences in the mean number of taste buds per papilla in the different age groups, nor was there a difference in taste bud diameter, which averaged 50 m across the groups. The number of fungiform papillae that did not contain taste buds ranged from 9 to 22% and there was no decrease in taste buds per papilla density in older animals.

In a later study, it was reported that fungiform papillae were larger towards the back of the tongue in comparison to the anterior regions of 6 adult cats and 5 kittens (Robinson and Winkles, 1990). The number of taste buds on each fungiform papilla ranged from 1 to 37, larger papillae at the back of the tongue contained the most taste buds and taste buds were identified on all of the fungiform papillae examined. In neonates, the papillae were smaller and contained fewer taste buds in comparison to kittens, however the number and distribution of fungiform papillae of kittens was similar to that of adults.

1.2.2.2 Human Studies

Lalonde and Eglititis (1961) studied the epiglottis, larynx, laryngeal pharynx and soft palate of cadaver tissues of an 18-hour-old newborn which contained circumvallate papillae, fungiform papillae, foliate papillae and filiform papillae. A total of 2583 taste buds were reported based on taste bud counts at all of the sites. Heiderich (1906) also reported a similar number, finding, an estimated 2500 to 3000 taste buds in the oral tissues of 0 to 11 months newborn cadavers. Taste bud density of the foliate papillae of 115

20 Chapter 1

Japanese cadaver tongues aged from newborn to 90 years old were investigated by Mochizuki (1939). The samples were divided into three groups; juvenile period (birth to 20 years) maturity period (21 to 60 years) and old age (61 to 90 years). Since the average foliate papillae density reported in the juvenile group was slightly higher (12.1 taste buds/ foliate papilla) than the mature (11.1 taste buds/ foliate papilla) and old age groups (10.3 taste buds/ foliate papillae), it was concluded that human taste bud density decreases with age.

Later, Arvidson (1979) examined 22 human cadavers aged from 2 days to 90 years old. Eight cadavers were babies aged 2 days to 6 months, and the others were adults aged 18 to 90 years. A range of 0 to 27 taste buds per fungiform papilla, with a mean of 0.97 taste buds per fungiform papilla was found for babies and a mean of 1.73 taste buds per fungiform papilla for adults. Recently, Segovia et al., (2002) compared taste bud and fungiform papillae densities in 20, 8 to 9 year old males and 20, 18 to 35 year old male adults. The data indicated that children had a significantly higher fungiform papillae density and taste bud density on the tip of the tongue than adults. Shahbake (2002) confirmed the results of the children in a study of 40 children (20 male and 20 female) from the same age group. In addition, it was reported that children with higher fungiform papillae and taste pore densities perceived PROP stronger than children with lower papillae and taste pore densities.

1.2.3 Adult to Elderly

Numerous studies have searched for possible causes and solutions to taste loss in aging people. Some researchers have focused on the biological changes that occur in the mouth, such as enhanced taste abilities or sensory deficits in the tongue, while others have focused on the impact of external factors on the mouth such as smoking. In general, there are two major views as regards the effects of aging on taste papillae and taste pore density, (1)

21 Chapter 1 the densities change with age (2) no changes occur. The following is a short summary of what it is known about taste bud loss with ageing.

Schieber (1992) suggested that the decline of taste could be related to the number of taste buds that a person loses on the surface of their tongue when they reach the seventh decade and cited several studies (Arey et al., 1935; Mochizuki 1937; Moses et al., 1967) that have reported humans lose 20 to 60% of their taste buds after this age. Hendricks et al., (1988), also noted that the number of taste buds on the tongue stays constant until the age of fifty when their numbers begin to decline and cause a decrease in sensitivity to salt and sweet substances Others reported that the average number of taste buds declines dramatically from 208 to 88 buds per circumvallate papilla when a person reaches the period 74 to 85 years (Arey et al., 1935). Researchers have speculated about whether or not a person's taste bud density changes as they age, causing gradual changes in the ability to recognize certain flavours of foods. Miller (1988a) for example, reported that taste bud density does not really diminish with age, but rather it stays at an equal level based on a person's health. Others, (Mistretta 1984; Bradley et al., 1985) reported that there was no significant difference in numbers of taste buds as a function of age in any type of lingual papilla. Cohen and Vig (1976) have explained the decreased taste bud density as being due to tongue growth, since the tongue approximately doubles in area from 4 to 18 years and growth continues beyond this age.

In summary, it appears that there are differences in papillae and taste bud densities from infants to adults, however, there are no indications when maturity is reached in humans. Thus, one of the aims of this thesis is to investigate whether the differences in fungiform papillae density stabilize by 12 years of age, by counting the papillae of 7 to 12 year old children and adults (Chapter 3, Part I).

22 Chapter 1

1.3 Methods Used to Study Peripheral Gustatory Anatomy

Taste papilla, taste pore and taste bud densities have been quantified in humans using several methods. One of the earlier methods involved use of the human cadaver tongue, however, procedures were later developed for live humans. Below are descriptions of the majority of the procedures that have been used to achieve these measures.

1.3.1 Taste Papillae

1.3.1.1 Magnifying Glass

One of the first methods used to count fungiform papilla on the tongue of live humans employed a magnifying glass. Moses et al., (1967) used an illuminated magnifying glass (X12) to count fungiform papillae in 200 healthy humans aged from 5 to 55 years old. The magnifying glass was fixed at a distance of 2 cm from a 1 X 1 cm frame. The frame was placed on five areas of the anterior tongue; one on each lateral aspect of the tongue; one on each side about 1 cm posteriorly from the tip and one at the same level in the midline. The number of fungiform papillae reported was 31 to 72 fungiform papillae/ cm2. The results were similar to later reports of Miller and Reedy (1990 a, b) who used a video-microscopy procedure with live humans and reported a range of 22 to 74 fungiform papillae/ cm2.

Although the magnifying glass produced similar results to that obtained with the video-microscope it has many limitations. In particular, no images can be captured with this method therefore (1) it is not possible to accurately compare results between subjects, or (2) map fungiform papillae on regions of the tongue, (3) it will also produce a lot of variability due to human error, because it relies on the skill of the investigator to identify the fungiform papillae and (4) reproducibility of the results is very difficult and time

23 Chapter 1 consuming as described by Moses et al., (1967) who counted the fungiform papillae of each subject on 10 separate occasions to confirm their results. In addition, with the low magnification of the magnifying glass, smaller papillae may not be identified and this method would not be suitable to identify or count taste pores.

1.3.1.2 Light Microscopy

Light microscopy in combination with staining procedures is another earlier method that was used to quantify and identify taste papillae and their taste buds (Miller 1988b). The tongue tissue is usually placed in formaldehyde in phosphate buffer for several days and the papillae dissected out from the anterior of the tongue, rinsed in water, dehydrated in graded ethanol and embedded in paraffin (Miller 1986). Serial sections which are usually longitudinal, are cut and stained with hematoxylin and eosin (H&E). The H&E stains create a contrast between the taste papillae and surrounding epithelial so that the papillae and their buds are stained lighter than the epithelial tissues (Figure 1.10), allowing their number and location to be recorded and analysed (Miller 1986).

TASTE PORE

Figure 1.10: Serial section of fungiform papilla, using H&E staining to show a taste pore using light microscopy (Gartner and Hiatt 2000)

24 Chapter 1

For example, Miller (1986, 1987) examined 1 cm2 areas from the tip of the tongue and mid-lateral regions in 10 human cadaver tongues aged from 22 to 80 years old, using light microscopy and frozen sections. The mean number of fungiform papillae reported were, 24.5 and 8.25 papillae/ cm2 in the tip and mid regions respectively. Cheng and Robinson (1991) used a similar method to study 26 human cadaver tongues (aged 17 to 92 years old) obtained within 12 to 24 hours of death. Using the Ponceau S red (pH 2.5) staining method they reported a mean of 195 in the first 2 cm of the tongue. Although, the procedures used in these studies to identify fungiform papillae were quantitative, they are invasive and not suitable for use with live humans.

1.3.1.3 Video-microscopy

A less invasive procedure is the use of video-microscopy to count fungiform papillae and taste pores on the human tongue (Miller and Reedy 1988). The video-microscopy unit consists of a dissecting microscope fitted with a video camera that records the magnified images of the tongue surface. Images are stored on a videocassette recorder and single frame images examined and saved with an image processor in a computer. The fungiform papillae are mapped and analysed at a higher magnification to identify taste pore density. The procedure used by Miller and Reedy (1988) involved the area of interest of the tongue being stained with 0.5% methylene blue.

In the method developed by Miller and Reedy (1990a) the subjects were seated on a chair, their tongues stained with the blue dye, following which they positioned their tongue in a tongue chamber, which was covered with a plastic cover slip for 2 to 3 minute intervals. For analysis, a map of all fungiform papillae on the anterior 2 to 3 cm of the tongue tip was constructed. Images of individual papillae were recorded from a small region (0.5 to 1.5 cm2) at higher magnification and diagrams were made to show the location of each papilla on the tongue tip and the number and location of individual taste pores on each papilla in a smaller tongue region. The average surface area

25 Chapter 1 mapped was 4.4 cm2 with an average of 41.1 fungiform papillae/ cm2 for a group of 12 subjects aged 17 to 33 years.

The latter findings were confirmed by the same authors (Miller and Reedy 1990b). The video-microscopy procedure was also used in combination with psychophysical procedures (Zuniga et al., 1993) and it was reported that individual differences of taste sensitivity on the anterior tongue are partly due to variation in the density of fungiform papilla and taste buds. The mean density of subjects aged 16 to 43 years was 8.2 papillae/ 43 mm2. More recently the video-microscopy procedure was modified for use with children (Segovia et al., 2002; Shahbake 2002). Segovia et al., (2002), quantified the number and size of fungiform papillae and pores in two areas (Figure 1.11) of the tongue that had been shown to have higher sensitivity to sucrose in children than adults (Stein et al., 1994). The subjects in the Segovia et al., (2002) study were 20, 8 to 9 year old males and 20, 18 to 30 year old male adults. Video-microscopy and NIH image software, were used with customized tongue templates (Stein et al., 1994; Segovia et al., 2002), and a red food dye was used to define the equivalent tongue location across the 40 subjects. The taste pores were stained with methylene blue. Unlike the procedure of Miller and Reedy (1990 a, b), the tongue was not held in a tongue chamber because the children were not able to keep their tongue still for a sufficient time to allow satisfactory images to be obtained. The fungiform papillae were counted in two areas; Area 1 was located in the tip of anterior tongue and Area 2 located in the mid region of the anterior tongue (Figure 1.11A). The mean fungiform papillae densities of Areas 1 and 2 were 8.9 and 3.5 per 9 mm2 in adults and 11.3 and 5.1 per 9 mm2 in children, respectively. Later Shahbake 2002 reported a similar fungiform density in Area 1 for 8 to 9 year olds (9.88 fungiform papillae/ 9 mm2).

26 Chapter 1

(A) (B) (C)

Figure 1.11: (A) and (B) show the 9 mm2 area investigated by Segovia et al., (2002) and Shahbake (2002), respectively. (C) Shows the 6 mm diameter area analysed by Shahbake (2002)

Shahbake (2002) adopted a similar procedure to investigate the relationship between the taste sensitivity of 40 children aged from 8 to 9 years with their papillae density. Minor changes were made to the procedure used by Segovia et al., (2002) and only Area 1 was measured (Figure 1.11B). The main change was that children were able to see the magnified image of their tongue on a TV screen sited in front of them which, to a great extent reduced tongue movement (Figure 1.12A and B). Shahbake (2002) also counted fungiform papillae on the tip of the tongue with the use of the 6 mm diameter filter paper (0.28 cm2 area) (Figure 1.11C). The above procedure was similar to Fast et al., (2002) who used a 6 mm diameter circle soaked in blue food dye to stain the tip of the tongue. The blue food dye method was only valuable for counting fungiform papillae as the dye only covers the surface of the tongue and does not stain the microvilli.

27 Chapter 1

The Subject

(A) (B) Figure 1.12: (A) The video-microscopy configuration. (B) Shows the position of the TV and subject in the amended video-microscopy procedure which reduced tongue movement in children (Shahbake 2002)

In summary, the video-microscopy method has been designed for use with adults and children and is a satisfactory non-invasive method for counting fungiform papillae. However, one of the limitations of the method is that it is not portable and therefore not suitable for use at the bedside of patients in an outpatient facility, or in locations such as schools outside the laboratory. Accordingly, one of the aims of this thesis is to develop a non-invasive and portable method for quantifying the number and size of fungiform papillae on the human tongue that can be used in remote locations.

1.3.1.4 Cell Counters

Another simple method for counting fungiform papillae density on the anterior regions of the tongue is the hand-held cell counter. Tepper and Nurse (1997) used this procedure with 75 adults aged 18 to 51 years old. The tip region (area = 0.39 cm2) of the tongue was stained with blue food dye to create a contrast with surrounding tissue, and the tongue stabilized with a wooden tongue depressor. Fungiform papillae density ranged from 47.2 to 75.8 fungiform papillae/ cm2, which was similar to the report by Miller and Reedy 1990 (a, b) using video-microscopy. Later Yackinous and Guinard (2002) used a cell counter to quantify fungiform papillae of 180 adults (69 males and

28 Chapter 1

116 females) aged 17 to 36 year olds. However the fungiform papillae density was higher than reported by Tepper and Nurse (1997), ranging from 116 to 165 fungiform papillae/ cm2. It was reported that these discrepancies could have been due to conducting the count in different regions of the tongue. For example, fungiform papilla density is higher in the anterior tongue in comparison with posterior regions. While the cell counter is a quick and simple method to use for counting fungiform papillae, it has the same limitations as the magnifying glass as regards reproducibility of the results, which are difficult and time consuming. In addition, no images are captured, fungiform papillae are identified at low magnification and it is not suitable for taste pore identification.

1.3.1.5 Contact Endoscopy

An endoscope is an illuminated optical instrument for visualizing the interior of a body cavity or organ. The fibre optic endoscope is very flexible, making it possible to reach inaccessible areas. Endoscopy has been used to view parts of the tongue that are difficult to reach and allow viewing of papillae physiology including the papillae vascular network or taste pores on the surface of the tongue at high magnifications. The contact endoscope has been used in combination with a xenon light source to view the surface of lingual papillae (Figure 1.13A) (Just et al., 2006). In that study the desired area was stained with methylene blue and the endoscope placed on the stained surface (Figure 1.13B) and single images or a video sequence recorded and analysed later.

29 Chapter 1

(B)

(A)

Figure 1.13: (A) Contact endoscopy configuration. (B) The contact endoscope on a stained tongue surface (Just et al., 2006)

Negoro et al., (2004) used the contact endoscopy method in combination with video microscopy to investigate the correlation between the condition of fungiform papilla and taste function in adults. The video microscope was used at a magnification of X50 to observe an entire papilla and the contact endoscope with a magnification of X60 was used to observe the blood vessels of a papilla. Blood vessels have been reported to become atrophic (Figure 1.14) and decrease in numbers in taste disorders (Umemoto et al., 1999 as cited in Negoro et al., 2004). The latter study qualitatively showed that the number of papillae with abundant blood vessels tended to be less on the tongues of subjects with taste disorders than on the normal tongue. The study classified fungiform papillae into 4 different types and blood vessels into five types. A summary of the different papillae types and blood vessels is given in Figure 1.15.

30 Chapter 1

(A) (B)

(C) (D) Figure 1.14: (A) Image of smooth egg-shaped fungiform papillae of a healthy participant with the use of a video microscope and (B) showing abundant straight blood vessels running parallel to each other in a fungiform papilla using contact endoscopy. (C) Image of flat and irregular fungiform papillae of a participant (video microscope) suffering chorda tympani injury and (D) shows the atrophic, irregular and less prevalent blood vessels (contact endoscopy) (Negoro et al., 2004)

It was concluded that video microscopy has the advantage of enabling the observation of many papillae in a large area of the tongue, but has the disadvantage of not allowing a clear observation of the vascular network of the papillae that can be seen with contact endoscopy.

31 Chapter 1

Figure 1.15: The classification of fungiform papillae and blood vessels (Negoro et al., 2004)

Contact endoscopy has also been used to compare the fungiform papillae density of patients suffering from middle ear disease with a group of healthy subjects (Just et al., 2006). In the study there were 14 patients from 14 to 61 years and 32 healthy subjects aged 7 to 71 years (mean 39.9  17.1 years). Measurement of papillae density was conducted in four different sections; tip of the tongue on the left and right side and the tongue edge on the left and right sides. No significant differences were found between the left and right sides on the tip or the tongue edges, with fungiform papillae density being 28.3 and 27.1 papillae/ cm2 on the left and right anterior tongue, respectively, and 15.6 and 15.7 papillae/ cm2 on the left and right tongue edges, respectively. However, there was a significant difference between fungiform papillae density on the tip and on the edge of the tongue and the younger subjects had a higher fungiform papillae density than older subjects regardless of the region. In other words ‘age’ was one reason for the decrease in papillae density. The latter anatomical data were also correlated with gustatory sensitivity measurements obtained by using an

32 Chapter 1 electrogustometry and ‘taste strips’ test kit which contained four tastants (sucrose, citric acid, sodium chloride and quinine). Overall, the patients exhibited a significant decrease in taste function and a significant decrease of fungiform papillae density on the side of the tongue that corresponded to the ear where surgery had been performed.

In summary, the contact endoscopy procedure appears to be an accurate and non-invasive method for viewing images of fungiform papillae and their vascular network on the surface of the tongue. However, to use this method for counting fungiform papillae is time-consuming because large sections of the tongue cannot be viewed at the same time and the device is not portable. Additionally, since contact endoscopy requires the tip of the endoscope to make contact with the surface of the tongue it would not be a suitable procedure to use with patients that have ulceration of the tongue due to illness or medical intervention. Contact endoscopy therefore, does not provide the necessary portability and fully non-invasive features required for the research described in this thesis.

33 Chapter 1

1.3.1.6 Summary of Findings Using Different Methods

Number of Age Reference Method Papillae density/cm2 subjects (years) Moses et al., 1967 MG* 31-72 200 5-55 Miller 1986 LM* 2.4-80 10 CT* 22-80 Miller and Reedy 1988 VM* 39.7 N/A N/A Kullaa-Mikkonen et al., 1987 EM* 14 tip, 9 edge, 3.6 mid region 127 20-65 Miller and Reedy 1990a VM 22.1-73.6 12 17-33 Miller and Reedy 1990b VM 22.1-74.3 16 17-29 Cheng and Robinson 1991 LM 171-253 (first 2 cm of AT*) 26 CT 17-92 Zuniga et al., 1993 VM 2.3-83.7 84 16-43 Tepper and Nurse 1997 CC* 47-76 75 18-51 Yackinous and Guinard 2002 CC 116-165 180 17-36 Segovia et al., 2002 VM 98.9 Adult, 125.5 Children 40 18-30, 8-9 Shahbake 2002 VM 110 40 8-9 Just et al., 2006 CE* 15.6-28.3 32 7-71

Table1.1: Summary of fungiform papillae density obtained using different procedures *MG: Magnifying glass LM: Light microscopy VM: Video microscopy EM: Electron microscopy CC: Cell counter CE: Contact endoscopy AT: Anterior tongue CT: Cadaver tongue

Table 1.1 is summary of the fungiform papillae densities using various procedures. There are differences in densities between and within different procedures which could be due to quantifying papillae in different regions of the tongue and the use of different age groups. Such areas can vary substantially as demonstrated by Stein et al., (1994) and Segovia et al., (2002). All of the above methods have only quantified taste papillae in small regions of the tongue especially the anterior tongue. The regions chosen were often arbitrary with the main reason being that higher fungiform papillae numbers occur on the anterior tongue in comparison to posterior regions.

On the other hand, quantification of a larger area of the tongue using some of the methods would be very difficult. For example, video-microscopy requires a frame-by-frame analysis of various sections of the tongue and requires a large amount of time for capturing and analysing images. Thus, one of the

34 Chapter 1 main aims of this thesis is to develop a rapid and efficient technique, which allows the study of a large part of the receptive field (i.e. anterior tongue for fungiform papillae) in sufficient numbers of participants of different ages, to allow a more valid assessment of the anatomical development of the peripheral gustatory system. None of the above methods can achieve this in non-laboratory environments as required in this thesis.

1.3.2 Taste Buds

1.3.2.1 Light Microscopy

Arvidson (1979) quantified fungiform papillae and their taste buds in 22 cadaver tongues (aged 2 days to 90 years) with the light microscope and the H&E staining procedure. Serial sections of 182 fungiform papillae indicated that taste bud density varied among papillae and between individuals. The number of taste buds in a single papilla ranged from 0 to 27, 63% of fungiform papilla contained no taste buds and 26% had 1 to 3 taste buds. Later, Miller (1986) reported a mean of 116 and 25.2 taste buds/ cm2 in the tip and mid-regions of human cadaver tongues, while the respective mean numbers of taste buds per papilla were 3.8 and 2.6. Subjects with the highest taste bud densities on the tip also had the highest densities in the mid-region and the highest number of taste buds per papilla. The above finding was supported by Miller (1987) and Cheng and Robinson (1991) who, used similar procedures to the latter study.

Light microscopy has also been used in live humans. Arvidson and Friberg (1980) compared the number of taste qualities recognized with the number of taste buds per papilla in a total of 110 fungiform papillae of 31 subjects aged from 18 to 35 years old. Individual papilla were stimulated with chemicals and taste buds counted by removing a single papilla with a scalpel and embedding it in paraffin for light microscopy. A total of 195 taste buds were reported in the 110 papillae examined (mean= 1.8), however, no taste buds

35 Chapter 1 were found in 56% of the papillae and these papillae also gave no taste response. The number of taste buds in the papilla varied from 1 to 15.

1.3.2.2 Electron Microscopy (EM) and Scanning Electron Microscopy (SEM)

Paran et al., (1975) used electron microscopy with excised human papillae to observe the different taste bud cell types (type I, II and III). Three to 5 papillae were removed from 11 healthy volunteers who received general anaesthesia for reasons other than papilla excision. Papillae with adequate epithelial margins were excised with a Hayes-Martin biopsy punch, which is a technique used for obtaining samples from the oral cavity and then fixed in buffer. Only the papillae from 3 subjects were satisfactorily preserved and only 1 to 3 papillae were examined in each subject. Two to 5 taste buds per papilla were found which was similar to the findings of Arvidson (1976), who studied the surface topography of human fungiform papillae by scanning electron microscopy. Arvidson (1976) removed 2 to 3 fungiform papillae from dental patients aged 20 to 50 years under anaesthesia and fixed the papillae in buffer. Taste pores were reported in 50% of the fungiform papillae and up to 3 pores in the dorsal surface were observed in single papillae. The average pore opening was reported to be 5 to 7m.

Later Kullaa-Mikkonen et al., (1987) used electron microscopy to count taste bud density in 85 live human tongues. In this instance the papillae were dissected from the tongue and fixed. Each specimen was then cut into two sections, one was processed for electron microscopy and the other for light microscopy. On average 1 to 5 taste buds were found on a single papilla. The diameter of the taste pores averaged 16 m, which was considerably higher than that reported by Arvidson (1976). The study also reported that 51 to 57% of the papillae did not have a taste bud. Although excision of fungiform papillae and EM and SEM are very accurate methods for identifying and quantifying taste buds of fungiform papillae, these procedures are invasive and not suitable for use with live humans, especially children.

36 Chapter 1

1.3.2.3 Video-microscopy

Miller and Reedy (1988) counted taste buds of 29 fungiform papillae of a human cadaver tongue using a video-microscope and reported that 10 of the fungiform papillae had no visible pores and the other 19 papillae had an average of 6.7 pores with a range of 1 to 14. The average tongue area studied was 1.09 cm2 and contained an average of 46.5 papillae. Taste pore density ranged from 36 to 511 pores/ cm2 with an average of 193.

Miller and Reedy (1990b) used video microscopy to measure the density of fungiform papillae and taste pores of 16 live humans in the first 2 to 3 cm of the anterior tongue and found that 91% of the papillae contained pores. The number of pores/ papilla ranged from 0 to 26 (mean= 4.56) and 8.8% of the papillae did not contain pores. The density of pores on the anterior tongue ranged from 36 to 582 pores/ cm2 with a mean of 254 pores/ cm2. The participants were divided into two groups with high and low taste pore densities. The higher density group had a mean of 374 pores/ cm2 with a range of 212 to 582 pores/ cm2, whilst the lower density group had a mean of 135 with a range of 36 to 173 pores/ cm2. Similarly, Zuniga et al., (1993) reported a mean taste pore density per subject of 25.2 pores/ papilla with range of 4 to 118. The number of taste pores per papilla was 3.1 with a range of 0 to 11.

As regards children, Segovia et al., (2002) reported 6.5 and 6.1 taste pores/ papilla in Areas 1 and 2 (Figure 1.11A), respectively, for 8 to 9 year olds children compared to 5.2 and 5.6 pores/ papilla, respectively, for adults. The mean pore density for the two areas was reported as 73.3 and 29.6 pores/ 9mm2 for children and 44.6 and 19.0 pores/ 9 mm2 for adults. Thus in both Areas 1 and 2, the 8 to 9 year olds had significantly higher papillae densities than adults. The data from the children were later confirmed by Shahbake (2002) who reported 76.03 pores/ 9 mm2 on the anterior tongue, which corresponded with Area 1 of the latter report (Figure 1.11B).

37 Chapter 1

1.3.2.4 Summary of Findings Using Different Methods

%p with Tb* or Tp* Number of Age Reference Method Tb/p* 0 Tb or density/cm2 subjects (years) Tp Arvidson 1979 LM* - 0-27 63 22 0-90

Paren et al., EM* - 2-5 - 11 - 1975

Arvidson 1976 SEM* - 3 50 - 20-50

Arvidson and LM 1-15 range 1.8 56 31 18-35 Friberg 1980

Miller 1986 LM 25.2 mid,116 tip 2.6 mid, - 10 CT* 22-80 3.8 tip

Miller and VM* 193 6.7 Tp/p 34 - - Reedy 1988

Kullaa- EM 14 tip, 9 edge, 3.6 - 127 20-65 Mikkonen et mid region al., 1987 Miller and VM 254 4.56 8.8 16 17-29 Reedy 1990b Tp/p

Cheng and LM - Up to 3 - 26 CT 17-92 Robinson 1991

Zuniga et al., VM 118 3.1 Tp/p - 84 16-43 1993

Segovia et al., VM 44.6 Adult, 73.3/ - - 40 18-30, 8-9 2002 9mm2 Children

Shahbake VM 76.03/ 9mm2 - - 40 8-9 2002 Children

Table1.2: Summary of taste pore density obtained using different procedures *Tp: Taste pore Tb: Taste bud p: papilla LM: Light microscopy VM: Video microscopy EM: Electron microscopy SEM: Scanning Electron Microscopy CT: Cadaver tongue

Table 1.2 is a summary of previous studies on the quantification of taste pore (bud) density per papilla and per cm2. There are many inconsistencies in taste pore densities measured with different methods. Children have a higher taste pore density on the anterior tongue in comparison to adults and higher

38 Chapter 1 taste pore densities are found on live human tongues in comparison to cadaver tissues.

The light microscopy and histological procedures used for pore and bud measures are invasive methods, and are not suitable to use with live humans, and all of the above methods described are designed for use in laboratories. At the commencement of this thesis the video-microscopy and contact endoscopy techniques were the only two suitable methods to quantify fungiform papillae and their taste buds in live humans. However, these procedures are not designed for use outside the laboratory or with large groups of participants. In particular they are not portable, fast or cost effective methods to quantify fungiform papillae in a large groups of participants in/or outside the laboratory. Accordingly, to achieve the several aims of this thesis the initial goal (Chapter 2) was to develop a portable and rapid method for quantifying taste papilla in large numbers of live humans that can be used outside the laboratory.

39 Chapter 1

1.4 Behavioural and Psychophysical Studies of the Development of the Gustatory System

1.4.1 Newborn to Infant

The ability to distinguish between different tastants have been demonstrated by variety of physiological and behavioural responses of newborn humans that include heart rate (Crook and Lipsitt 1976), brain electrode activity (Fox 1985), bodily movements, facial expressions, sucking (Crook and Lipsitt 1976; Crook 1978) and tongue movements (Rosenstein and Oster 1988).

Preference for sweet tastants is shown by positive hedonic responses by newborns by increased sucking, tongue movement and heart rate (Desor et al., 1975; Beauchamp and Moran 1984). Furthermore, Rosenstein and Oster (1988) and Nicklaus et al., (2005) observed negative hedonic responses to sour and bitter tastants in comparison to water and sweet solutions, which manifested as behaviours including restlessness, disruption of sucking and negative facial expressions. In regards to salt perception Desor et al., (1975) and Ganchrow et al., (1983) reported that infants cannot discriminate between salty solutions and water, and Beachamp and Cowart (1984) suggested that infants may be insensitive to salt with sensitivity gradually increasing over the first 2 years of life. Similarly, others reported a “distaste” response to salt but no distinctive expression in infants and concluded that in infants the ability to distinguish salt is not completely developed (Rosenstien and Oster 1988).

40 Chapter 1

1.4.2 Adolescent to Adult

In general children are less sensitive to tastants than adults. For example, Hermel et al., (1970) showed that 7 to 10 year olds were less sensitive to sucrose than adults which was consistent with earlier research by Richter and Campbell (1940) and later James et al., (1997) who showed that children were less sensitive than adults when a whole mouth stimulation technique was used. The latter result was in contrast to the finding by Stein et al., (1994), who reported higher sweet sensitivity on the tip of the tongue for 8 to 9 year olds compared to adults. James et al., (1997) suggested the difference was due to children not being able to integrate neural information as efficiently as adults, which could indicate the taste system is immature in 8 to 9 year olds.

Human taste acuity has been reported to decrease with age. Studies have reported that both the detection threshold and recognition thresholds are elevated in older persons (Schiffman 1993; Murphy 1993). For example, Moore et al., (1982) studied 71 adults aged 20 to 88 years and found that sucrose taste detection thresholds increased gradually with ageing.

1.4.3 Gustatory Perception

Although taste is a physiological system, taste perception is affected by the senses of vision, taste and smell. Another factor that effects our taste perception is past experience. It is likely that people learn and become familiar with specific combinations of colors and tastes and these learned associations may alter our perceptions and create expectations about how a food or drink should smell and taste.

According to previous studies there is a strong acceptance of sweet taste in infancy and childhood (Desor et al., 1975; Mennelle et al., 2005) with preference levels decreasing to that of adults during late adolescence (Desor

41 Chapter 1 and Beauchamp 1987). Genetic variation (Bartoshuk 2000), and early experience (Beauchamp and Moran 1984) also play a role in establishing individual differences.

Besides taste perception variation between children and adults there are also individual differences in genetically determined sensitivity to certain tastes. From birth to old age the ability to taste compounds that contain an N-C = S group, such as phenylthiocarbamide (PTC) and its relative propylthiouracil (PROP) is evident in human populations (Blakeslee 1932; Kalmus 1976; Bartoshuk 1979). Although these chemicals taste bitter to some, others either cannot taste them or require high concentrations to recognize their presence. The degree of taste sensitivity has been shown in some studies to be associated with disliking/ liking of bitter and sweet tastes, and, it may have important long-term health implications related to dental caries, obesity and diabetics (Tepper 1998; Duffy et al., 2003; Rupesh and Nayak 2006; Tepper et al., 2014). For example, Rupesh and Nayak (2006) reported that children that were less sensitive to bitterness of PROP were “sweet-likers” and preferred strong tasting food products, while the majority of children with higher sensitivity to PROP were “sweet-dislikers” and preferred weak tasting food products.

Furthermore, the ability to taste PTC/ PROP has been widely used in taste sensitivity research. Bartoshuk et al., (1994) categorised adults into three groups based on their sensitivity to supra-threshold concentrations of PROP. The categories were non-tasters, who accounted for approximately 25% of the population, medium-tasters (50%) and super-tasters (25%) (Bartoshuk et al., 1994; Mennella et al., 2005; Tepper et al., 2014). The latter study also reported that taste papillae densities vary as a function of PROP-taster status, indicating that super-tasters have more papillae on the tip of the tongue as well as a greater overall density of taste pores than medium and non-tasters (Bartoshuk et al., 1994; Tepper and Nurse 1997). Later, Shahbake (2002) reported similar classifications for children (8 to 9 years old) with 25% of the 40 children categorised as non-tasters, 45% as medium-

42 Chapter 1 tasters and 30% as super-tasters and reported significant correlations fungiform papillae density and sensitivity to PROP.

In summary there are many studies that have shown differences in taste perception between adults and children and the basis for these differences could be due to incomplete development of children’s gustatory systems peripherally and/ or centrally.

43 Chapter 1

1.5 Diseases and Taste Disorders

In the previous sections the development of the peripheral gustatory system and various methods used to quantify taste papillae and their taste pores were briefly described. Since one aim of this thesis is to investigate the effect of disease, in particular renal disease on taste perception, in the following section a review is given of taste disorders arising from diseases and clinical treatments.

The disorders of taste are classified as “geusias” and are divided into four groups; ageusia is the inability to detect taste, hypogeusia is a decreased ability to detect taste, dysgeusia is distorted perception of taste, and hypergeusia is increased sensitivity to some or all tastes (Nelson 1998). Taste loss is widespread and associated with a variety of illnesses. Poor oral hygiene is a leading cause of hypogeusia and cacogeusia (detection of a normal taste as foul and unpleasant taste) and viral, bacterial, fungal and parasitic infections may lead to taste disturbances (Bartoshuk et al., 1983). Furthermore, oral cavity and mucosal disorders including oral infections, inflammation and radiation-induced mucositis can impair taste sensation. Impaired taste function may also be caused by lesions at a variety of sites including the oral mucosa, taste buds, unmyelinated nerves or cranial nerves to the brain stem (Cheng 2007; Vissink et al., 2003a, b). It has also been reported that malignancies/ tumours of the head and neck are associated with decreased appetite and inability to detect flavours (Mossman 1983; Zheng et al., 2002; Sandow et al., 2006).

Other conditions in which taste loss may occur include xerostomia (dry mouth), Sjögren syndrome (inflammation of the salivary glands resulting in a dry mouth) (Henkin et al., 1972), liver (Deems and Friedmen 1988; Reiter et al., 2006) and kidney disorders (Astback et al., 1999), depression (Osaki et al., 1996; Kiewe et al., 2004) and epilepsy (Baumann et al., 2004).

44 Chapter 1

Taste loss may also be caused by decreased levels of zinc (Blendis et al., 1981; Osaki et al., 1996), copper (Ciges and Morales 1980) and nickel (Shatzman and Henkin 1981) and hormonal fluctuations during menstruation and pregnancy. (Breen et al., 2006; Ochsenbein-Kolble et al., 2005). Other causes are Type I familial dysautonomia (i.e., Riley-Day syndrome) which results in severe hypogeusia or ageusia because of the absence of taste bud development (Smith et al., 1965; Gadoth et al., 1982), multiple sclerosis, facial paralysis and thalamic lesions (Malauiya and Ramu 1981; Combarros et al., 1994). Lichen planus (Petruzzi et al., 2002), aglycogeusia (Henkin and Shallenberger 1970), erythema multiform and geographic tongue (Lozada- Nur et al., 1992; Abdollahi and Radfar 2003) are other causes.

Various types of therapy can induce taste loss. For example, drugs usually induce a temporary effect, which diminishes after they are discontinued. In chemotherapy there are number of medications that are associated with taste loss including methotrexate, dexamethasone and antihypertensives. Once the medications are discontinued the taste acuity of patients commonly return to normal (Petruzzi et al., 2002).

45 Chapter 1

1.5.1 Taste and Common Diseases in Children

Taste in children is vulnerable to a range of diseases and treatments consistently seen in paediatric clinics. The following is a summary of what is known on this topic.

1.5.1.1 Cystic Fibrosis

One of the diseases that is reported to alter taste perception is cystic fibrosis. This disease is also known as mucovoidosis, or mucoviscidosis and is a hereditary disease affecting the exocrine (mucus) glands of the lungs, liver, pancreas, and intestines, causing progressive disability due to multisystem failure. Thick mucus production results in frequent lung infections and diminished secretion of pancreatic enzymes is the main cause of poor growth, greasy stools, and deficiency in fat-soluble vitamins (A, D, E and K). Henkin and Powell (1962) reported that cystic fibrosis patients had greater salt taste and olfactory acuity than clinical controls, whilst in contrast others (Wotman et al., 1964) reported hyposensitivity and lower sensitivities towards phenylthiocarbamide (PTC) (Manlapa et al., 1965). Among the patients 45.3% were classified as tasters and 54.7% as non-tasters, whereas their matched controls consisted of 70% of tasters and 30% non-tasters. Later Desor and Maller (1975) and Hertz et al., (1975) reported that identification of a salt solution was within the normal range for children with cystic fibrosis, which was recently confirmed by Laing et al., (2010).

46 Chapter 1

1.5.1.2 Diabetes Mellitus

Loss of taste is also a problem in diabetes mellitus. The disease is a multi- systemic disorder of carbohydrate, protein and fat metabolism. Diabetes mellitus may cause serious problems in the oral cavity such as gingivitis, periodontitis, candidiasis, glossitis, oral ulcerations, loss of taste sensations and tooth decay (Bell et al., 2000; Fowler et al., 2001). One of the reasons for oral complications is diabetic peripheral neuropathy, which can cause loss of sensations in the oral cavity and decreased salivation (Collin et al., 2000). Recently Gardiner et al., (2008) reported patients suffering from diabetes have problems with neurotrophic support, which is critical for the development of sensory receptors on the tongue. Therefore, a decrease in neurotrophic support may cause reductions in the number of taste papillae and taste bud density, which would likely lead to taste loss. Decreases in taste acuity have been reported by Terry and Segal (1947) who found an association between diabetes mellitus and the inability to taste PTC and reported that this loss increased with time after the onset of disease. The latter finding was also confirmed by others (Rao and Sisodia 1970; Bhatia et al., 1991), who reported that non-taster diabetics have poorer taste acuity for glucose than the taster diabetics. Lawson et al., (1979) and Bhatia et al., (1991) also reported significantly higher glucose thresholds in subjects with diabetes. In conclusion, diabetes may decrease the ability of a patient to recognise sweet tastants, which can affect perception of other tastants.

1.5.1.3 Otitis Media

Another disease that can affect taste sensitivity is otitis media, which is an infection or inflammation in the middle ear. The inflammation is caused by a build-up of fluid behind the eardrums (Sone et al., 2001). Taste can be affected because the major taste nerve, the chorda tympani, travels behind the tympanic membrane, through the middle ear on its path from the tongue to the brain stem (Bartoshuk et al., 1996). The chorda tympani innervates the

47 Chapter 1 taste buds of the anterior tongue and detects all four tastes (Sone et al., 2003). Thus, during otitis media, the chorda tympani can be damaged because it is exposed to the pathogens that characterise the disease (Bartoshuk et al., 1996). Also during middle ear surgery the chorda tympani can be subjected to surgical stress by stretch, injury, or dryness which may cause symptoms such as dysgeusia, hypogeusia, or ageusia (Sone et al., 2001; Sone et al., 2003; Nin et al., 2006). Bartoshuk et al., (1996) showed that subjects with otitis media had significantly lower taste bud counts than healthy subjects, whilst Gedliki et al., (2001) reported lesions of the chorda tympani in adult patients with otitis media.

Later, Landis et al., (2005) studied changes of gustatory function in adults due to otitis media and associated this disease with significant lower taste sensitivity to sweet and salt tastants on the anterior two thirds of the tongue on the side of the infected ear in comparison to the healthy side. However, no significant alteration in the perception of bitter and sour tastes was reported between the healthy and diseased side. Additionally, the latter investigators reported that there was a trend to greater impairment in relation to the severity of the inflammatory process. Overall, otitis media is reported to affect the taste sensitivity and the taste anatomy of individuals and these alterations are likely to influence their food preferences.

1.5.1.4 Radiation Therapy

Post-irradiation gustatory dysfunction is the taste loss that occurs with the administration of radiation therapy to the oral cavity, which may cause a decrease in a patient’s ability to eat causing inadequate nutrition, which can lead to severe weight loss (Bolze et al., 1982). The changes in taste thresholds for all tastants (sweet, sour, bitter, salty and umami), varies among treated patients (Mossman 1982). For example, lower sensitivity to bitter tastants was reported by a number of investigators (Bonanni and Perazzi 1965; Congar 1973; Mossman and Henkin 1978), however, the loss of salt, sour and sweet tastants was also found to vary considerably among

48 Chapter 1 patients (DeWys 1974). Recently, Yamashita et al., (2006b) reported that taste acuity declines during the 5th week after irradiation and improves by the 11th week after ceasing irradiation.

Another reason for taste loss during radiation therapy, whether it is temporary or permanent, has been linked to loss of taste buds of gustatory papillae. Thus, Conger (1973) reported that taste buds degenerate 6 to 7 days after irradiation. This was confirmed by Esses et al., (1988), who reported degeneration after 8 days and Yamashita (2006a) who reported degeneration by 7 to 9 days. Yamashita et al., (2006a) also reported that the causes of taste disorders are dependent on two factors. One factor is the disappearance of taste buds and the other is the dysfunction of the salivary glands. The latter results in dryness in the mouth leading to the inability of the taste substances to penetrate taste pores. Thus it was concluded that taste disorders induced by radiation are mainly caused by damage to the taste buds within the radiation field.

1.5.1.5 Renal Disease

Renal failure may result from many other diseases including diabetes, hypertension and polycystic kidney disease (Douma and Smit 2006; Sarafidis et al., 2008). Chronic renal failure can cause impaired taste acuity (Matson et al., 2003). Studies have reported poor sensitivity for all four common tastes (sweet, sour, salt and bitter) (Atkin-Thor et al., 1978). One of the causes for taste loss may be zinc deficiency (Matson et al., 2003). Chronic renal failure, also affects taste bud density and Astback et al., (1999), reported that patients with the disease had significantly fewer taste buds than controls. Reasons for taste bud loss were not given. Currently the study by Astback et al., (1999) is the only one that has reported a variation in taste bud counts in individuals with chronic renal failure. The study was conducted by removing papillae from the tongue surfaces of patients, which is an invasive method and is not suitable for use with children. Since a study of the effects of renal disease and children’s taste function is described in Chapter 5, a more

49 Chapter 1 comprehensive literature review on the effect of chronic renal failure on the gustatory system is presented in the Introduction of that Chapter.

1.6 Summary

Psychophysical and behavioural studies have demonstrated that there are differences in taste perception and food preferences between children and adults. Additionally, anatomical studies using various techniques including light microscopy, electron microscopy, video-microscopy and contact endoscopy, have reported that children have higher fungiform papillae and taste bud densities than adults. However, there are no studies indicating at what age the human gustatory system matures. To investigate this question, a large group of children and adults are required together with a rapid, portable and accurate procedure for measuring anatomical characteristics. At the commencement of this thesis no such method existed. Furthermore, there are many diseases that cause taste loss and only a few studies have been reported that assess children. According to the literature the majority of the methods developed to monitor taste loss are based on (1) animal models and (2) invasive methods that are not suitable for children. There are no studies in the literature that have monitored taste loss from the anatomical perspective in live humans that are non-invasive. The following therefore are the objectives of this thesis, which aims to create a better understanding of the development of the human gustatory system and better measurement of taste losses due to diseases and medical interventions.

50 Chapter 1

1.7 Aims of Thesis

The aims of this thesis are:

1) Develop a reliable and portable method for quantification of fungiform papillae on the human tongue (Chapter 2) (Shahbake et al., 2005).

2) Investigate the variation in fungiform papillae density on the whole anterior tongue during development from childhood to adulthood using the new technique (Chapter 3, Part I) (Correa et al., 2013).

3) Design a predictive model for children and adults to calculate fungiform papillae density of a larger area of the tongue using fungiform papillae counts of smaller areas (Chapter 3, Part II) (Correa et al., 2013).

4) Investigate the validity of previous methods that used small areas on the anterior tongue as representative of larger areas for calculation of fungiform papillae density on the anterior tongue (Chapter 3, Part III) (Correa et al., 2013).

5) Develop a new technique that eliminates the use of staining procedures during the quantification of human fungiform papillae (Chapter 4) (Correa et al., 2015).

6) Apply the new technique for counting papillae and the non-staining procedure to a clinical situation (Chapter 5) (Correa et al., 2015).

51 Chapter 2

Chapter 2

Development of a Rapid Method for Counting Fungiform Papillae

52 Chapter 2

2.1 Introduction

The methods commonly used to count taste papillae have been described in Chapter 1. As indicated in the Chapter, the magnifying glass (Moses et al., 1967) and light microscopy techniques (Miller 1988a) have been used to quantify fungiform papillae on the anterior tongue and both have the disadvantage that they are time-consuming. Others have used excision of taste papillae from the tongue surface of live humans (Paran et al., 1975), however this method is invasive and not suitable for use with children. Video- microscopy (Miller and Reedy 1988; Segovia et al., 2002; Shahbake 2002) and contact endoscopy methods (Negoro et al., 2004; Just et al., 2006), however, are two non-invasive techniques that could be used in vivo to count taste papillae. Both methods have been verified as a suitable procedure to quantify fungiform papillae, however, both methods, have a disadvantage of not being portable and require a substantial amount of time to analyse each image and quantify taste papillae or taste buds. Thus one of the objectives of this Chapter is to develop a portable procedure for counting fungiform papillae on the anterior human tongue. A candidate procedure is one that uses a digital camera.

The digital camera was considered to be a suitable instrument because it is portable and easy to assemble and produces images with high magnification and resolution. To develop and validate such a new procedure, it must be compared against an established technique. Accordingly, the present Chapter compares the data obtained using a digital camera and a video- microscope by measuring fungiform papilla density on the tongue of adults and children. This Chapter have been published in a peer review journal; Brain Research (Shahbake et al., 2005) (Appendix E, Paper 1).

53 Chapter 2

2.2 Aims

The aims of Chapter 2 are (1) develop a new portable and rapid method for counting fungiform papillae and (2) validate the new method by comparing the data obtained with that using video-microscopy.

2.3 Methods

2.3.1 Participants

The participants were 9 children aged 8-9 years [mean of 8.7  0.17 (S.E.) years] and 7 adults aged 25-38 years [mean of 31.0  1.54 (S.E.) years], who were recruited from local suburbs and were assessed in a University microscopy laboratory. Both children and adults were chosen because the video-microscopy method has been used for these age groups at this laboratory (Segovia et al., 2002; Shahbake 2002).

The procedures of the study were demonstrated to all participants and in regards to children the procedure was also shown to their parents or guardian, who gave written consent prior to the experiment and were present for the duration of the experiment. All participants received compensation (movie tickets and confectionaries) at the end of the study. The study was approved by the University of Western Sydney Human Research Ethics Committee (approval number HEC02/188).

54 Chapter 2

2.3.2 Equipment

The following are descriptions of the digital camera and video-microscopy methods.

2.3.2.1 Video-microscopy Configuration

The arrangement for video-microscopy was adopted from a previous study conducted by Shahbake (2002). The unit consisted of a SZ6045TR Olympus microscope with a SZ CTV attachment for a CCD 72S Dage MI camera. A Sony video recorder (SLV-815 HQ) was connected to this camera and to a 68 cm Panasonic TV, to view and focus the section of the tongue under the microscope. To view and enhance the black and white images obtained by the video-microscopy set up, a Euromex fiber optic source (Model EX-1) was used (Figure 2.1).

Video recorder

Fibre optic light source

Graphic printer

Video camera

Microscope

Figure 2.1: The video-microscopy configuration (Shahbake 2002)

55 Chapter 2

2.3.2.2 Digital Camera Configuration

The digital camera used was a Nikon Coolpix 4500, which produced digital images of 4.0 megapixels. At the time this project commenced this was the highest resolution camera available that contained a rotating head. The camera was equipped with a macro light attachment (Macro Cool-light SL-1) to enhance the tongue image when the “macro” option was chosen in the menu setting (Figure 2.2).

Macro light attached to the rotating head LCD screen to view the images

Figure 2.2: The Nikon Coolpix 4500 camera with the macro light and rotating head

2.3.3 Experimental Procedure

The experiment was conducted over two sessions, which were performed on the same day to ensure accurate measurement of the same stained tongue area. In session 1 the digital camera method was used, because of its rapid procedure, followed 5-10 minutes later in session 2 by the video-microscopy procedure.

56 Chapter 2

2.3.3.1 Staining Procedure

Prior to taking images, the tongue was stained with blue food dye (Robert’s Brilliant Blue FCF133) to create a contrast between the surrounding tissues. The dye only covered the surface of the papillae and did not stain the microvilli of the taste papillae. Therefore, this staining procedure is only suitable for counting taste papillae not for counting taste pores. Initially, a participant rinsed their mouth with deionized water (Milli-Ro-6 Plus System, conductivity 0.9 S) then dried their tongue with a filter paper. The tongue was then stained with blue food dye, which was prepared as a paste (Fast et al., 2002; Shahbake 2002). Using a steel forceps a circular 6mm diameter (0.28 cm2 circle) filter paper (Whatman’s No. 1) was dipped in the paste and placed on the left anterior tongue for 3 seconds. The tongue was dried with a clean filter paper to prevent the dye from spreading and to reduce glare when the photographic images were obtained (Figure 2.3).

57 Chapter 2

Forceps

(1) (2)

0.28 cm2 circle soaked in Before staining, the tongue blue food dye and placed on was dried with filter paper to the tongue prevent the dye from spreading

(3) Filter paper was (4) removed Stained area of Tongue was dried with filter the tongue paper to prevent blue dye spreading

(5)

Stained area ready for videomicroscopy and digital camera procedures

Figure 2.3: Diagram of the staining procedure with blue food dye

58 Chapter 2

2.3.3.2 Digital Camera Procedure

A participant was asked to sit on a chair and to support their head by placing their arms on a table such that their chin protruded forward (Figure 2.4A). Such an arrangement minimized head movement and reduced tongue movement dramatically. The participant then protruded their tongue and held it steady with their lips. A 10 x 3 mm wide piece of filter paper (1 cm scale) was placed on the right side of the anterior tongue, providing a scale to calculate the magnification of each image (Figure 2.4B). The use of a scale was very important because every image captured had a different magnification depending on the distance of the camera from the tongue. Following this, 3 to 5 images of the stained area were captured with the digital camera and later analysed.

(A) (B)

Figure 2.4: (A) Demonstrating the seating position of the participant (B) the tongue showing the stained area and the 1 cm filter paper scale on the other side of the tongue

59 Chapter 2

2.3.3.3 Video-microscopy Procedure

Participants were seated and their neck supported with a neck pillow (Figure 1.12A). The participants protruded their tongue and they could see the enlarged image of their tongue in the TV screen in front of them which reduced the head movement considerably (Figure 1.12B) (Shahbake 2002). The area stained prior to using the digital camera procedure was then recorded for 30 to 45 seconds at a low magnification (X3). The stained area was magnified with the optical lens of the microscope and recorded at X30 to X50 magnification for a duration of between 5 to 12 minutes. The recordings were from different angles to ensure the entire surface of the stained area was captured.

2.3.4 Image Capture Analysis

The images captured from both the digital camera and video-microscopy procedures were then analysed separately and the fungiform papillae counted. All the images were numbered by a 3-digit code and analysed randomly for both methods. The examiner was blinded to the image source for individuals within a method and did not know which subject was being assessed. The 0.28 cm2 circle stained area was remeasured in both methods for every participant as some of the dye spread beyond the edges of the filter paper. Each fungiform papilla was then numbered on a print of the image to ensure only the fungiform papillae in the 0.28 cm2 circle were included in the measurements (Figure 2.5 and 2.6). Papillae on the border of perimeter of the 0.28 cm2 circle were included for analysis if most of a papilla was inside the perimeter.

60 Chapter 2

2.3.4.1 Digital Camera Image Capture

The images captured were downloaded to a computer (Toshiba Satellite PRO6100) using the USB port and the images analysed with Adobe Photoshop 7.0. The files were opened using this program then the “zoom” option was chosen to magnify the images. Based on the 1 cm scale the tongue magnifications were calculated. The image magnification was calculated and recorded when it was suitable to view the fungiform papillae in the screen and all the fungiform papillae in the desired area were visible. The 1 cm filter paper was measured with “ruler” option in the “view” menu. The magnification used in the camera ranged from X9.0 to X15.5, however, if a fungiform papilla was too small to identify the magnification was increased to X20 to X24.8 to correctly identify the papilla (Figure 2.5). After the data about a papilla were quantified and recorded they were statistically analysed for any differences between the digital camera and video-microscopy methods.

Figure 2.5: Image of the tongue captured by the digital camera. Arrows show fungiform papillae in the stained area at X21 magnification

61 Chapter 2

2.3.4.2 Video-microscopy Image Capture

After the video recording, the tapes were viewed and still images produced from VCR images. The images were then captured on the computer to count fungiform papillae. These analyses were performed on a G3 Macintosh computer, which was installed with a Scion LG3 frame grabber card using NIH Image software V1.61. The images captured were in black and white as per previous established video-microscopy procedures (Miller and Reedy 1990a; Segovia et al., 2002; Shahbake 2002).

To capture images the “start capturing” command was chosen from the “special” menu. The capture rate was 30 frames per second allowing clear pictures to be obtained. To store pictures on the computer the images were highlighted with the mouse and then from the “stack” menu the “capture frames” option chosen. The videos were viewed using a slow motion frame- by-frame feature and the chosen pictures were captured on the computer by clicking the mouse button. Once the images were captured they were enhanced. From the processing menu the “smooth” filter option was chosen to reduce noise and soften the images and the “sharp” filter was then chosen to increase the contrast. The magnifications of the images were X30 (Figure 2.6).

Figure 2.6: Image of the tongue captured by the video-microscopy procedure. Arrows show fungiform papillae in the stained area at X30 magnification

62 Chapter 2

2.3.5 Fungiform Papilla Identification

Fungiform papillae have been described as red spots with a mushroom-like elevation on the surface of the tongue (Negoro et al., 2004). Some fungiform papillae have been described as short and cylindrical and similar to conical papillae, others as short and wide-necked (Miller 1995; Segovia et al., 2002; Shahbake 2002). Fungiform papillae on the tip of the tongue can appear as, two papillae joined together (“double papillae”) and others contained a hair like projection (fila) (Kullaa-Mikkonen and Sorvari 1985; Miller 1995; Segovia et al., 2002). In the present study all of these papilla shapes were observed, except the papillae containing fila. Similar observations were made by Segovia et al., (2002) who also reported such papillae were not seen in children, only in adults (Figure 2.7).

The digital camera produced colour images whereas black and white images were produced by the video-microscopy method. In both colour and black and white images fungiform papillae were identified according to the above criteria.

The blue food dye only covered the surface of the fungiform papillae, so that the fungiform papillae were seen as small, blue, round mushroom-like structures in the digital camera images and grey round structures in the video-microscopy images. The surrounding filiform papillae were seen as light blue, conical and flatter structures than fungiform papillae in the digital camera images and a lighter grey shade in the video-microscopy images.

63 Chapter 2

(A) (B) (C)

(D) (E) (F)

Figure 2.7: Examples of different shaped fungiform papillae using the video-microscopy method: (A) “double papilla”, (B) flat papillae, (C) short and wide necked mushroom shaped papilla (Shahbake 2002). The digital camera method: (D) “double papilla” and flat papillae, (E) raised and conical shaped papilla, (F) Short and wide necked papilla

2.3.6 Statistical Analysis

For each participant the fungiform papillae were counted using both the digital camera and the video-microscopy methods. A 2X2 analysis of variance (ANOVA) was performed to check for any significant difference between factors [fungiform papillae count (dependent variable) with the digital camera method versus video-microscopy method] and within factors (adults versus children) (Appendix A, Table A2).

64 Chapter 2

2.4 Results

2.4.1 Fungiform Papilla Counts with the Digital Camera and Video-microscopy Methods

The mean fungiform papilla density in the 0.28 cm2 stained area using the digital camera method was 10.29  1.10 (S.E.) in adults and 15.00  0.99 (S.E.) in children. The ranges for adults and children were 6 to 14 and 10 to 19 papillae, respectively. The overall (adult + children) mean fungiform papillae density was 12.94  0.93 (S.E.)/ 0.28 cm2 (Appendix A, Table A1).

The results for the video-microscopy method indicated that the mean fungiform papilla density in the 0.28 cm2 area was 9.71  1.17 (S.E.) in adults and 13.78  1.02 (S.E.) in children. The fungiform papillae ranges were 6 to 14 and 10 to 19 fungiform papillae/ 0.28 cm2 for adult and children, respectively. The overall (adult + children) mean fungiform papillae/ 0.28 cm2 was 12.00  0.91 (S.E.) (Appendix A, Table A1).

2.4.2 Comparison of Papillae Counts Obtained Using the Digital Camera and Video-microscopy Methods

The ANOVA showed there was no significant main effect between the two methods of counting (F(1,28)= 0.697, p= 0.411). However it indicated there was a significant main effect between the adults and children due to children having the higher density, (F(1,28)= 16.695, p< 0.001). Additionally, there was a significant interaction between factors (Method X Age) (F(3,28)= 53.19, p= 0.003). Since there were significant interactions between factors an interaction graph was plotted to examine where these differences appear (Appendix A, Table A2). Such a graph is a line plot in which each point indicates the mean of a dependent variable (fungiform papillae density) at one level of a factor (Method). The level of a second factor (Age) is used to construct separate lines. For two or more factors, parallel lines indicate that there is no interaction between factors, which means that only levels of one factor could be investigated. In Figure 2.8 even though, there are only two

65 Chapter 2 points representing the two counting methods in children and adults, it illustrates that the lines are not parallel confirming there is an interaction between the two factors (Method X Age) (F(3,28)= 53.19, p= 0.003).

Although the sample groups for this section were small the findings were similar to that of the previous study by Segovia et al., (2002) who measured the fungiform papilla density of 20 adults (18-30 years) and 20 children (8-9 years) using the video-microscopy method and reported that children had a significantly higher papilla density than adults (p<0.01).

Figure 2.8: Interaction plot showing interactions between the two factors (Method versus Age)

66 Chapter 2

2.4.3 Individual Subject Differences in Fungiform Papillae Counts Between the Two Methods

In addition to comparing fungiform papillae counts between adults and children with the two methods, individual counts were also examined between methods. Although in the overall results the differences were averaged out and there were no statistical differences between the two methods (F(1,28)= 0.697, p= 0.411), there were large differences between the two methods as regards the counts obtained for certain individuals. For example, there was a significant difference in fungiform papillae counts between the two methods for participants 4 and 8 in the children’s group

(Figure 2.9A) and 12 and 15 in the adult group (F(1,7)= 15.00, p= 0.008) (Figure 2.9B) (Appendix A, Table A3).

67 Chapter 2

20

18

16

2 2 14 cm

12 0.28 Digital camera 10 Video-microscope 8 Count/

6 Papillae count/28.27mm 4 Papillae 2 0 123456789 (A) Children

16

14

2 2

cm 12

0.28 10

Digital camera 8 Video-microscope Count/

6

Papillae count/28.27 mm 4 Papillae

2

0 (B) 11 12 13 14 15 16 17 Adults

Figure 2.9: Fungiform papillae counts for individuals using the digital camera and the video- microscope for (A) Children and (B) Adults. Arrows show large differences in counts between the two methods for four participants.

68 Chapter 2

The fungiform papillae images of the four participants who exhibited large differences were compared across the two methods to identify reasons for the large variation between the two methods. The digital camera method produced a higher fungiform papillae count than the video-microscopy method in all four participants. The main reason for lower papillae counts in the video-microscopy images was associated with defining certain papillae. The images captured by the video-microscopy method were in black and white, therefore, the contrast created between the stained area and surrounding filiform papillae were different shades of grey making the small fungiform papilla difficult to identify. In comparison, it was easier to identify papillae using the digital camera because there was a greater contrast between the blue food dye and pink surface of the tongue (Figure 2.10).

Digital camera Videomicroscopy

1 2 10 10 2 1 ? 9 11 3 9 ? 8 16 3 8 4 7 12 7 ? 5 6 15 4 14 6 13 5 ? 17 11

Figure 2.10: Comparison of images from the two photographic methods of the same area showing the visibility of the fungiform papillae of participant 4

For example, in Figure 2.10 in the digital camera image 17 fungiform papillae were identified compared to only 11 in the video-microscopy image. The other 6 papillae were not identified because they were not visible. Overall, undercounting using the video-microscopy occurred with 8 participants and with 5 participants there was an over-count because what was identified as papillae in the video-microscope images were recognized as surrounding filiform papillae in the digital camera images and equal counts occurred with 3 participants (Figure 2.9).

69 Chapter 2

If the two images of both methods are compared together it is possible to identify these differences but if the methods are used individually it is possible to undercount the fungiform papillae using video-microscopy due to the angle of the image and the contrast of the stained area with surrounding filiform papillae. These differences create a potentially larger error in fungiform papillae identification using video-microscopy.

2.4.4 Summary of Findings

Overall, there were no significant differences in mean counts of fungiform papillae of the two subject groups between the digital camera and the video- microscopy methods. However, there were significant differences in papillae counts between the adult and children groups within the individual methods. Furthermore, it was easier to identify fungiform papillae in the colour images captured with the digital camera than with the video-microscope method, improving the ability of the assessor to identify the papillae.

70 Chapter 2

2.5 Discussion

The main objective of this Chapter was to investigate whether the use of a digital camera to count fungiform papillae would produce similar results to those obtained with a video-microscope. According to the statistical analysis there was no significant difference in the counts between the two methods. The fungiform papillae density was 36.40 and 53.10 fungiform papillae/ cm2 in adult and children respectively using the digital camera method and 34.35 and 48.74 fungiform papillae/ cm2 with the video-microscopy procedure.

There were differences in fungiform papillae counts between the present study and previous reports. For example the papillae counts of the adults were lower than previous reports by Miller and Reedy (1990a, b) and Segovia et al., (2002) who reported 41.1, 43.8, 99 (Area 1) and 38 (Area 2) fungiform papillae/ cm2 respectively using the video-microscope method (Figure1.11A). However the papillae counts of the adults were higher than Zuniga et al., (1993) study, who also used the video-microscopy method and reported 19 fungiform papillae/ cm2. As regards to children, the fungiform papillae counts of the present study were also lower than previous reports. Segovia reported 125 and 56 fungiform papillae/ cm2 for children in Areas 1 and 2 (Figure1.11A).

The most likely reason for the differences in fungiform papillae counts between studies is the location of the tongue that was chosen for conducting the counts in addition to the age of the groups, as it has been reported that children have significantly higher papillae counts on the tip of the tongue than adults (Segovia et al., 2002). For example Miller and Reedy (1990 a) study was based on the on the anterior 2 to 3 cm of the tongue tip whereas, Segovia et al., (2002) counted papillae in two 9 mm2 areas (Area 1 and 2). Area 1 was located in the tip of anterior tongue and Area 2 located in the mid region of the anterior tongue.

71 Chapter 2

The use of different techniques for counting papillae is another reason for variations in papillae numbers being reported. The papillae counts of this study were also lower than previous studies that used a different counting method, such as the cell counter (Tepper and Nurse 1997; Yackinous and Guinard 2002). The cell counter method, for example, has limitations including the fact that no images of the papillae are produced or saved, therefore reproducibility and verification is difficult and increases the experimental error.

In addition to the video-microscopy method being more difficult to identify papillae, another disadvantage of the method was the amount of time it required for recording, capturing and analysing the images before the papillae count occurred. A period of 20 to 30 minutes was required for adjustment of the video-microscopy equipment depending on the participant and the recording of the stained area at low magnification, which was suitable for counting papillae. Afterwards, still images were captured from recordings and the papillae were counted, resulting in a time period of 45 to 60 minutes to analyse the stained areas of a tongue.

In contrast, using the digital camera, from the time the participant was seated until the images of the stained areas were captured required 4 to 7 minutes depending on the participant’s tongue movement. Overall, the total time to capture and count the fungiform papillae of each participant was between 20 to 25 minutes, which was a saving of approximately 60% in time for each participant.

The most important advantage of the digital camera, was it’s portability. Importantly, the present study demonstrated that the camera and associated equipment could be assembled quickly for use in schools, Outpatient Clinics and at a patient’s bedside in contract to the bulky equipment used with the video-microscopy. Since the digital camera method was less time-consuming it is also be suitable for conducting studies with large groups of people or in conjunction with psychophysical studies such as that of Bartoshuk et al., (1994) who classified individuals based on their taste sensitivity to PROP in

72 Chapter 2 relation to fungiform papillae density, or Essick et al., (2003) who used fungiform papillae counts in conjunction with taste sensitivity tests to measure the taste acuity of participants.

In conclusion, the present study has demonstrated that the digital camera provides a valid method for counting fungiform papillae on the human tongue (Shahbake et al., 2005). The digital camera is utilised in the next chapter (Chapter 3) to study fungiform papillae density from childhood to adulthood using a large number of subjects.

73 Chapter 3

Chapter 3

Quantification of Fungiform Papillae Density During Human Development

74 Chapter 3

3.1 General Introduction

Childhood obesity is a major health problem in Australia and other countries and its prevalence is on the increase (Lobstein et al., 2004). Obesity is associated with disorders that include neurological, cardiovascular, respiratory, gastrointestinal, endocrine and psychological problems (Must and Strauss 1999). Many studies have reported that the daily consumption of fruit and vegetables amongst children is below the recommended levels (Liu et al., 2000; Magarey et al., 2001; Wardle et al., 2003). Whether the taste perceptions or food preferences of children are sufficiently different from adults to be part of the cause is unknown, but preferences and taste perception do change throughout life (Koehler and Leonhaeuser 2008).

For example, a reason for low consumption of fruits and vegetables in children may be due to rejection of bitter and sour tastes (Koivistto and Sjoden 1996; Akella et al., 1997). The ability to taste the bitter substance PTC/ PROP is present in young children and declines slowly with age (Whissell-Buchey 1990). Also 8-9 year old children can be classified into three PROP-taster groups (non-tasters, medium-tasters and super-tasters) (Shahbake 2002). As regards sweet and salt taste it has been reported that children prefer higher concentrations of sweet (Desor et al., 1975; Cowart 1981) and salt than adults (Desor et al., 1975; Beauchamp et al., 1985; Beauchamp et al., 1990; Mennella et al., 2003) and they are more sensitive to sweet in smaller regions of the anterior tongue in comparison to adults (Stein et al., 1994).

One reason for variation in taste preferences could be anatomical differences between children and adults. For example, Segovia et al., (2002) reported that children (8-9 year olds males) have more fungiform papillae than adults on the tip of the tongue. Thus, children may perceive the strength or quality of tastes differently to adults. Since children by 8-9 years are still poorer at defining the 4 tastes their preference for foods may be affected because of their immaturity (James et al., 1997). In addition Shahbake (2002) reported

75 Chapter 3 that 8-9 year old children with higher fungiform papillae and taste pore densities, perceived PROP to be stronger than children with lower papillae and taste pore density on a small area on the tip of the tongue. However, there are no studies that have reported at what age maturity is reached as regards to numbers of taste papillae. Thus Chapter 3 has several goals which are aimed at resolving when differences in papillae density between adults and children disappear and how best to count papillae on the anterior tongue. This Chapter have been published in a peer review journal; Chemical Senses (Correa et al., 2013) (Appendix E, Paper 2).

3.2 Aims

The aims of Chapter 3 are (1) in Part I, to determine at what age fungiform papillae density reaches adult levels, (2) in Part II, to investigate whether fungiform papillae counts of small regions on the anterior tongue are suitable predictors of the fungiform papillae density of the whole anterior tongue, and (3) in Part III, check whether previous methods correctly indicated the number of fungiform papillae on the anterior tongue of adults.

The region of the anterior tongue examined in this Chapter was the first 2 cm from the tip of the tongue because the majority of fungiform papillae are located on the tip (Miller and Reedy 1990a) and extend to about the first 2 cm of the tongue (Temple et al., 2002) while, it is sparsely distributed over the posterior area. Temple et al., (2002) demonstrated that the anterior tongue ceases to grow at 8-10 years of age, whereas the posterior region continues to grow until 15-16 years of age. A critical finding in Temple et al., (2002) study was that the area of the anterior tongue in 8-10 year olds and adults is the same. Hence, this equivalence in size allows direct comparison of papillae density in any part of the anterior tongue of adults and children of this age.

76 Chapter 3

3.3 General Methods

3.3.1 Participants

A total of 115 participants were recruited for Parts I, II and III. The participants were 29 children aged 7-8 years [mean= 7.78  0.47 (S.E.) years], 23 aged 9-10 years [mean= 9.68  0.51 (S.E.)], 33 aged 11-12 years [mean= 11.60  0.49 (S.E.)] and 30 adults aged 20-24 years [21.63  0.95 (S.E.)]. All participants’ data were utilised for all three Parts of this Chapter It should be noted different participants were used in each Chapter. The adults were University students and the children were recruited from local primary schools. Children of these age groups were chosen because it has been reported that children as young as 7 years old can perform basic psychophysical tasks and they can be trained to follow procedures similar to those required in the present study (Stein et al., 1994; Laing et al., 2008). The procedures were described in the information sheet that was taken home by the children to their parents or guardian to sign. All participants received compensation (movie tickets and confectionaries) at the conclusion of their participation. The children were assessed at their school and the adults at the University microscopy laboratory. The study was approved by the University of Western Sydney Human Research Ethics Committee (approval number HEC02/188).

3.3.2 Equipment and Experimental Procedure

The equipment utilized in this Chapter for all three Parts was the digital camera and associated instrumentation as described in Chapter 2. The experimental procedure was also similar to that described in Chapter 2, except for the staining procedure which in Parts I, II and III involved staining the whole left anterior region of the tongue.

77 Chapter 3

3.3.2.1 Staining Procedure

Participants rinsed their mouth with deionized water (Milli-Ro-6 Plus System, conductivity 0.9 S) and their tongue was dried with a filter paper by the experimenter prior to commencing measurements (Figure 3.1 (1)). A piece of filter paper (Whatman’s No. 1) that contained a blue food dye (Robert’s Brilliant Blue FCF133) was placed on the anterior left side of the tongue for 3 seconds (Figure 3.1 (2)). On removal of the filter paper the tongue was dried with a clean filter paper (Figure 3.1 (4)). The area stained was the most anterior 3-4 cm of the dorsal tongue. This length was used since it covered more than the 2 cm “demarcation line” which defined the anterior and posterior region of the tongue in adults and children (Temple et al., 2002). As in Chapter 2 (section 2.3.3.2), head movement was minimized by the participant supporting their head by placing their arms on a table and holding their head with their hands such that their chin protruded forward. The participants were asked to protrude their tongue as far as possible because a larger section of the anterior tongue was being stained in comparison to Chapter 2. They were also asked to hold their tongue steady with their lips. A 10 x 3 mm wide piece of filter paper placed on the right side of the anterior tongue provided a scale to calculate the magnification of each image (Figure 3.1 (5)). Following this, 3 to 7 images of the stained area were captured with the camera. Images were also taken from different angles to ensure all of the stained areas were in the images captured.

78 Chapter 3

(1) Tongue was dried with filter paper prior to staining

(2) Filter paper soaked with (3) The filter paper was removed blue food dye was placed on the left anterior tongue for 3 seconds.

(4) The stained area was (5) A 1 cm scale was placed on dried with a filter paper to the right side of the tongue after prevent the dye from which the tongue was ready for spreading photographing

Figure 3.1: Stepwise diagram of the staining procedure

79 Chapter 3

Part I

Fungiform Papillae Density and Their Distribution Patterns in Children and Adults

80 Chapter 3

3.4 Introduction

According to reports there are differences in papillae and taste bud densities from infants to adults (Mochizukie 1939; Schieber 1992; Segovia et al., 2002; Just et al., 2006), however, none of these studies indicates when maturity is reached as regards to the number of taste papillae and taste buds.

One of the major aims of the present research was to examine a large group of participants from childhood to adulthood and discover at what age anatomical differences as regards fungiform papillae disappear. Since, a portable, fast and cost effective method was required to achieve this aim the digital camera method in Chapter 2 was used for counting fungiform papillae on the human tongue.

3.5 Aims

The objective of Part I was to investigate when the number and distribution of papillae on the anterior tongue reaches adult levels. The participants were divided into four groups, including, children aged from 7 to 12 years and a group of adults. Fungiform papillae were counted in the anterior tongue using the digital camera method to determine at what age adult levels are attained.

81 Chapter 3

3.6 Methods

3.6.1 Image Analysis

Prior to image analysis, a computerized grid was superimposed on top of the image using the “grid” option in the “view” menu (Figure 3.2 (1)). The size of each grid was 0.5 X 0.5 cm. A worksheet was used to record the data from each grid and per cm (Appendix B, Table B1). The midline and each cm of the stained area (cm 1 and 2) from the tongue tip were marked using the horizontal and vertical “guide” option in the “view” menu (Figure 3.2 (1)) with cm 1 being the closest cm to the tongue tip. After the grids and guides were superimposed on the desired image, the number of papillae in each region were analysed using a “zoom” option in the “view” menu in the Adobe Photoshop 7.0 program. The total fungiform papillae were counted on the first 2 cm of the anterior tongue and per cm intervals then statistically analysed (Figure 3.2 (2)). The magnification used for the image analysis ranged between X14.12 to X23.73 (mean= X18.43).

All images were identified by a 4-digit number and randomised, so that the identity of the participants and their groups were unknown to the experimenter. Using the grids the fungiform papillae were counted from left to right. This was important for recounting, data reproducibility and when the stained area was subdivided (Part II).

82 Chapter 3

cm 2 Midline

cm 1

(1) Lines were drawn to mark the midline and divide the stained area into two 1 cm Computerised grid intervals. A computerised grid was superimposed on the tongue image for a more accurate count of papillae.

(2) The fungiform papillae in each part of the grid square were counted from left to right and recorded per grid. Figure 3.2: Diagram demonstrating preparation of the images before counting fungiform papillae

83 Chapter 3

3.6.2 Statistical Analysis

Prior to data analysis the participants were divided into four groups: (1) 7-8, (2) 9-10, (3) 11-12 year olds and (4) adults. The analysis conducted was based on the first 2 cm of the tongue because two previous studies (Miller and Reedy 1990a; Temple et al., 2002) reported that there were very few fungiform papillae beyond the first 2 cm.

For each participant the following measurements were made: (1) the total fungiform papillae count in the first 2 cm of the tongue and (2) the fungiform papillae count per cm intervals for the first 2 cm of the stained area (Appendix B, Tables B2-B5). A 2 X 4 (Sex X Group) between-participants ANOVA was performed to examine whether significant differences existed between the fungiform papillae densities of all four groups and between males and females across the groups (Appendix B, Table B6). In addition, two 2 X 4 (Sex X Group) between-participants ANOVAs were performed to determine if there were significant differences (1) within the fungiform papillae density of the first and second cm intervals of the tongue and (2) within the fungiform papillae density of all groups in these sub-regions and (3) between males and females of all groups (Appendix B, Tables B8 and B10).

84 Chapter 3

3.7 Results

3.7.1 Overall Variation of Fungiform Papillae Density Between Age Groups

Group Total FP* FP range 1st cm FP 2nd cm FP Total FP M/F*  (S.E.) counts  (S.E.) counts  (S.E.) counts  (S.E.) Male Female 7-8 years 162.24  3.74 116-195 109.93  2.63 52.31  2.90 159.08  6.41 164.47  4.59 9-10 years 146.09  4.69 75-186 93.87 3.09 52.22  2.43 140.07  5.21 157.38  8.33 11-12 years 150.79  5.27 85-217 93.39  4.24 57.40 2.81 149.71  6.27 151.94  8.81 Adults 140.20  5.50 93-222 91.30  4.07 48.90  3.01 134.11  8.25 141.95  7.11

Table 3.1: Counts of total fungiform papillae and per cm intervals of all groups. * FP = Fungiform papillae, M= male and F= female

Table 3.1 shows the total fungiform papillae counts of the anterior 2 cm stained area and the fungiform papillae counts in each cm interval for all groups. The first cm in the table indicates the cm closest to the tongue tip. It also contains the total fungiform papillae density of each group sub-divided into male and female participants (Appendix B, Tables B2-B5).

A 2 X 4 (Sex X Group) ANOVA analysis was performed amongst all age groups to determine if there were significant differences in the total fungiform papillae counts. Overall there was a significant difference between groups

(F(3,107) = 3.238, p= 0.025) but, no significant difference between males and

females (F(1,107) = 0.033, p= 0.855). Also there were no significant Sex X

Groups interactions (F(7,107) = 0.974, p= 0.408) (Appendix B, Table B6). Post- hoc Tukey LSD paired comparisons (p= 0.05) showed that 7-8 year olds had significantly more papillae than adults (p=0.010) but there were no significant differences between any other groups (Appendix B, Table B7).

85 Chapter 3

3.7.2 Variation of Fungiform Papillae Density Per Cm Intervals Between Age Groups

The data were further analysed per cm interval to investigate whether there were any significant differences amongst groups in the various areas of the anterior tongue.

3.7.2.1 Variation of Fungiform Papillae Density in the First Cm of the Anterior Tongue

There was a significant difference in the fungiform papillae counts in the first cm closest to the tip across groups (F(3,107) = 5.282, p=0.002). Post-hoc Tukey LSD paired comparisons (p= 0.05) showed there were significant differences between 7-8, 9-10, 11-12 year olds and adults (p<0.05) (Appendix B, Table B8), however, there were no significant differences between 9-10, 11-12 year olds and adults (p>0.05) (Appendix B, Table B9).

As regards gender, there were no significant differences between males and females within each age group (F(1,107) = 0.021, p=0.885). Interestingly, the analysis showed an overall significant Sex X Groups interaction (F(7,107) = 2.993, p=0.007) (Appendix B, Table B8). Thus an interaction graph was plotted to investigate where these differences appear (Figure 3.3). The plot shows two different trends for males and females. In males the fungiform papillae density progressively decreased from 7-8 year olds to adults, while in females there was a density decrease from 7-8 to 9-10 year olds where it remained stable to adulthood.

86 Chapter 3

Figure 3.3: Interaction plot showing variation between males and females in different age groups

3.7.2.2 Variation of Fungiform Papillae in the Second Cm of the Anterior Tongue

A 2 X 4 ANOVA performed on the fungiform papillae density counts in the second interval indicated there were no significant differences amongst groups (F(3,107) = 1.288, p= 0.282), or between male and females (F(1,107) =

0.249, p= 0.619). There were also no significant interactions (F(7,107) = 0.804, p= 0.586) between Sex X Groups (Appendix B, Table B10).

However, there seemed to be a variation in fungiform papillae shapes amongst groups. Papillae were identified as small and round structures in 7-8 year olds (Figure 3.4A). The shapes of the fungiform papillae in 9-10 year olds were similar to 7-8 year olds but they appeared slightly larger in size (Figure 3.4B). In 11-12 year olds and adult groups the fungiform papillae

87 Chapter 3

were irregular in shape (Figure 3.4 C and D). All of the shapes for the different age groups were similar to previous findings (Segovia et al., 2002).

(A) (B)

(C) (D)

Figure 3.4: The Tongue of (A) 7-8 year old, (B) 9-10 year old, (C) 11-12 year old and (D) an adult. Arrows show fungiform papillae

88 Chapter 3

3.7.3 Distribution Patterns of Fungiform Papillae in Children and Adults

Following the counts of fungiform papillae the distribution pattern of papillae of each individual was determined. In addition to fungiform papillae density differences with age, qualitatively different distribution patterns of papillae were found.

The distribution patterns in this study were divided into two types. Type A was defined as having clusters of closely packed fungiform papillae in the tip of the tongue and a rapid decrease of papillae numbers towards the posterior region of the tongue (Figure 3.5A). Type B contained high papillae numbers on the tip and had a more homogenous distribution, but there were no obvious clusters of papillae (Figure 3.5B). The patterns were achieved by drawing an outline line around a participants’ tongue image. A dot was superimposed for each fungiform papilla and the midline and 2 cm interval marked for consistency between participants (Figures 3.5 and 3.6).

(A) (B)

Figure 3.5: Examples of different distribution patterns showing drawing of tongues of two 8 year olds. The dotted line indicates the end of the second cm. (A) Type A, arrow showing cluster of fungiform papillae (B) Type B, evenly distributed fungiform papillae

89 Chapter 3

Overall, both types were observed in all groups of children with 72.4%, 69.6% and 72.7% of 7-8, 9-10 and 11-12 year olds categorized as Type A and, 27.6%, 30.4% and 27.3% of 7-8, 9-10 and 11-12 year olds categorized as Type B (Figure 3.7A-F), (Appendix B, Tables B2-B5). Figure 3.7 shows examples of Type A and B distribution patterns of 7-8, 9-10, 11-12 year olds and adults.

Type A and B distributions were also observed in adults. However, clusters were less compacted together in comparison to those of the children (Figure 3.6) and there were more Type B (63.3%) distributions in adults than Type A (36.7%) (Figure 3.7G-H). Interestingly, in adults, most of the Type A distributions occurred in females (81%) (Appendix B, Table B5). Overall no other type of papillae distribution was found in any group.

(A) (B)

Figure 3.6: Diagram of different distribution patterns in the adult group, the dotted line is an indication of the end of the second cm. (A) Type A, arrow showing cluster of fungiform papillae (B) Type B, evenly distributed fungiform papillae

90 Chapter 3

FP cluster

(A)

No FP cluster

(B)

Figure 3.7: Tongues of 7-8 year old children showing different papillae distribution patterns. (A) Type A, arrow indicating cluster of fungiform papillae (FP) and (B) Type B, arrow showing no cluster of FP; scale= 1 cm.

91 Chapter 3

FP cluster

(C)

No FP cluster

(D)

Figure 3.7 (Cont.): Tongues of 9-10 year old children showing different distribution patterns. (C) Type A, arrow indicating cluster of fungiform papillae (FP) and (D) Type B, arrow showing no cluster of FP; scale= 1 cm.

92 Chapter 3

FP cluster

(E)

No FP cluster

(F)

Figure 3.7 (Cont.): Tongues of 11-12 year old children showing different distribution patterns. (E) Type A, arrow indicating cluster of fungiform papillae (FP) and (F) Type B, arrow showing no cluster of FP; scale= 1 cm.

93 Chapter 3

FP cluster

(G)

No FP cluster

(H)

Figure 3.7 (Cont.): Tongues of adults showing different distribution patterns. (G) Type A, arrow indicating cluster of fungiform papillae (FP) and (H) Type B, arrow showing no cluster of FP; scale= 1 cm.

94 Chapter 3

3.7.4 Summary of Findings

The aim of Part I was to investigate at what age children attain similar numbers of fungiform papillae, distribution and growth of the papillae in the anterior tongue to adults. This was achieved and the results show that by 9- 10 years of age the number of fungiform papillae reached adult level and the distribution and density of fungiform papillae stabilized at 11-12 years of age.

Overall, the data indicated there was a significant decrease in fungiform papillae density in the first cm of the anterior tongue with age which reached adult level by 9-10 years of age. The fungiform papillae of 11-12 year olds and adults appeared irregular in shape in comparison to those of the younger children, who had smaller and rounder papillae. These data suggest maturity is reached as regards papillae numbers and shape by 11-12 years of age. Also, there were no significant differences of fungiform papillae density between males and females in any age group.

Finally, there were two distribution patterns of fungiform papillae observed in both children and adults. Type A contained a cluster of fungiform papillae in the tip of the tongue and a decrease of fungiform papillae towards the mid region of the tongue. However, in Type B the fungiform papillae were distributed more evenly. The majority of the children’s fungiform papillae distribution patterns were categorized as Type A whereas in adults the majority of the patterns were categorized as Type B.

95 Chapter 3

Part II

Predicting Fungiform Papillae Density

96 Chapter 3

3.8 Introduction

The most frequent location that has been used by others to count fungiform papillae is a small area on the left anterior tongue close to the tip of the tongue and mid-line (e.g. Tepper and Nurse 1997; Fast et al., 2002). For most people the tip region contains the highest density of fungiform papillae in comparison to the posterior regions and these are also easily accessible for measurement. Accordingly, the latter area has been arbitrarily used as an indicator of the fungiform papillae density on the anterior tongue (Fast et al., 2002), however, no study has demonstrated that one small region on the tip of the tongue is more suitable than another as an indicator of the fungiform papillae density on the whole anterior tongue.

3.9 Aims

The aim of Part II was to determine whether fungiform papillae density in one or more small regions on the anterior tongue are suitable predictors of the fungiform papillae density of the whole anterior tongue.

3.10 Methods

In Part I fungiform papillae density was counted in the first 2 cm of the tongue, then into two 1 cm regions within the 2 cm area. However the 1 cm intervals still represented large regions for counting fungiform papillae in studies with large groups of participants. Therefore, in Part II the two 1 cm long areas were subdivided into 8 areas using the images captured in Part I, and the papillae densities measured in each of the 8 areas (Figure 3.8). The data were then analysed to develop a model that indicated which area(s) was the best predictor(s) of the overall fungiform papillae density on the anterior tongue. The procedures followed for this section were the same as in Part I except for the image analysis.

97 Chapter 3

3.10.1 Image Analysis

As indicated, the images captured from Part I were analysed differently in Part II. Prior to counting the fungiform papillae, each 1 cm band was divided into 4 smaller sections using the “grid” option in the “view” menu, resulting in a total of 8 small areas (Figure 3.8). Areas 1 to 4 were in the first cm band and Areas 5 to 8 were in the second band. Fungiform papillae were counted in each of the 8 areas, recorded, analysed and compared with the total number of fungiform papillae in the two cm region that encompassed most of the anterior tongue. The length and width of each small area was also measured based on the 1 cm scale and the area of each of the 8 areas calculated.

7 8 2nd cm 5 6 3 4

1 2 1st cm

(A) (B)

Figure 3.8: (A) The image of the stained area was sub-divided into small areas. (B) A diagram of the tongue showing the numbering sequence of the 8 small areas

98 Chapter 3

3.10.2 Statistical Analysis

The analysis was based on the four age groups defined in Part I and it was investigated which of the 8 small areas was the best predictor(s) of the overall fungiform papillae density in the anterior tongue.

For each age group the following counts were made; total fungiform papillae density of the first 2 cm anterior tongue and fungiform papillae counts in each of the 8 small areas (Appendix B, Tables B11-B14). A stepwise multiple regression analysis (SPSS 12.0.1) was performed using the total number of fungiform papillae in the first 2 cm (dependent variable) anterior tongue and the counts in each of the 8 small areas (predictors).

For all age groups three models of predictor equations were calculated based on the statistical analysis of the 8 small areas versus the total number of fungiform papillae. The models for each age group are shown in Tables 3.2 to 3.5. The multiple regression technique identified which area(s) best predicted the total fungiform papillae density, and a model was developed from the statistical analysis. The equation was; (y= bA1 + bA2 + bA3 + ……..+ Constant) where b was the coefficient calculated from the regression analysis, AX represented the area number, and the constant was calculated from the regression analysis. For each group and the combined groups the best three-predictor equations were calculated (Tables 3.2 to 3.6).

3.11 Results

3.11.1 Predictor(s) of the Total Fungiform Papillae Count for All Age Groups

The regression performed between the total number of fungiform papillae in the first 2 cm of the anterior tongue and the papillae numbers per 8 small areas showed that for 7-8 year olds Area 1 (R2 = 0.44) was the best predictor of the overall total fungiform papillae indicating that fungiform papillae density

99 Chapter 3 of the anterior tongue in this age group can be predicted by counting the papillae in Area 1. The addition of Areas 5 and 7 to the model increased the prediction to 69% (Table 3.2).

The strongest predictor of papillae density for 9-10 year olds was also Area 1 (R2= 0.45) and inclusion of Area 5 increased R2 to 62% (Table 3.3). Interestingly, the strongest predictor for 11-12 year olds was Area 5 (R2= 0.52) and inclusion of Area 4 in the equation increased R2 to 72% (Table 3.4). Similarly, with adults the strongest predictor was Area 5 (R2= 0.46) and addition of Area 1 increased R2 to 60% (Table 3.5). The data were also analysed by including the data from all groups and this indicated the best predictor of the total fungiform papillae count was Area 1 (R2= 0.33) whilst inclusion of Area 5 increased R2 to 56% (Table 3.6).

Model Predictor Equation R2 Area (s) 1 y= (2.281A1 + 60.85) 0.44 (44%) 1 2 y= (2.259A1 + 2.048A5 + 31.898) 0.56 (56%) 1 and 5 3 y= (2.009A1 + 2.990A5 + 1.904A7 +7.429 0.69 (69%) 1, 5 and 7

Table 3.2: The predictor models of total fungiform papillae of the anterior tongue of 7-8 year olds.

Model Predictor Equation R2 Area (s) 1 y= (1.701A1 + 84.563) 0.45 (45%) 1 2 y= (1.779A1 + 2.422A5 + 50.787) 0.62 (62%) 1 and 5 3 y= (1.806A1 + 1.916A5 + 2.210A7 + 9.959) 0.71 (71%) 1, 5 and 7

Table 3.3: The predictor models of total fungiform papillae of the anterior tongue of 9-10 year olds.

Model Predictor Equation R2 Area (s) 1 y= (3.067A5 + 100.215) 0.52 (52%) 5 2 y= (2.522A5 + 2.516A4 + 58.28) 0.72 (72%) 5 and 4 3 y= (2.123A5 + 2.477A4 + 2.141A2 + 25.932) 0.81 (81%) 5, 4 and 2

Table 3.4: The predictor models of total fungiform papillae of the anterior tongue of 11-12 year olds.

100 Chapter 3

Model Predictor Equation R2 Area (s) 1 y= (3.190A5 + 92.569) 0.46 (46%) 5 2 y= (2.195A5 + 0.865A1 + 61.313) 0.60 (60%) 5 and 1 3 y= (2.894A5 + 0.638A1 + 1.203A10 + 48.321) 0.69 (69%) 5, 1 and 2

Table 3.5: The predictor models of total fungiform papillae of the anterior tongue of adults.

Model Predictor Equation R2 Area (s) 1 y= (2.373A1 + 128.266) 0.33 (33%) 1 2 y= (2.302A1 + 1.069A5 + 107.112) 0.56 (56%) 1 and 5 3 y= (1.769A1 + 0.976A5 + 1.344A7 + 97.537) 0.60 (60%) 1, 5 and 7

Table 3.6: The overall predictor models of total fungiform papillae of the anterior tongue of all age groups.

45

40

35

30

25 7‐8 yrs papillae/area 20 9‐10 yrs 11‐12 yrs 15 Adults

Fungiform 10

5

0 12345678 Area number

Figure 3.9: Fungiform papillae counts in each of the 8 small areas in each of the groups

Figure 3.9 shows that the density decreased from Areas 1 to 8 whilst Area 1 at the tip of the tongue contained the highest fungiform papillae density for all groups.

101 Chapter 3

3.11.2 Summary of Findings

Overall, the aim of Part II was to determine whether fungiform papillae counts in one or more small regions of the anterior tongue are suitable predictors of the papillae density of the whole anterior tongue. This was achieved and the results showed that Area 1 was the best predictor of overall papillae density for 7-10 year olds, whereas for 11-12 year olds and adults the best predictor was Area 5. Equivalent predictor location in 11-12 year olds and adults is another indicator suggesting that that maturity is reached by 11- 12 years of age.

102 Chapter 3

Part III

A Comparative Study of Fungiform Papillae Counts of the Anterior Tongue

103 Chapter 3

3.12 Introduction

Part II showed that depending on the age group, Areas 1 and 5 were the strongest predictors of the total fungiform papillae density on the anterior tongue. Even though these areas were small (Area 1: 0.35 cm2 and Area 5: 0.43 cm2), previous studies (Fast et al., 2002; Yackinous and Guinard 2002) used even smaller areas (0.28 cm2 circle) to count fungiform papillae and assumed the value obtained was representative of the anterior tongue (Fast et al., 2002). The locations chosen in previous studies were generally close to the tip of the tongue because of the high papilla density. For example, a 6 mm diameter circle of filter papers (0.28 cm2 circle) was the marker in several studies (Tepper and Nurse 1997; Fast et al., 2002; Yackinous and Guinard 2002). The filter paper was soaked in blue food dye and placed on the tip of the tongue and the fungiform papillae in the stained area counted.

However, no study has demonstrated that the area(s) chosen was a suitable predictor of the papillae density of the anterior tongue. Accordingly, the aim of Part III was to check whether the data on papillae densities in small regions reported by others correctly predicted densities for the whole anterior tongue.

3.13 Aims

The aims of Part III were to (1) determine whether previous measures of papillae density using a 0.28 cm2 circle correctly predicted the total number of fungiform papillae and 2) check whether other area(s) (i.e. Areas 1 and 5) were better predicators of fungiform papillae counts in children and adults than those used in previous studies.

104 Chapter 3

3.14 Methods

In Part III, fungiform papillae were quantified in two locations designated as Area X and Area Y. Area X corresponded to mainly Area 1 of Part II and a small portion of Area 3, and Area Y corresponded mainly to Area 5 and a small section of Area 7 (Figure 3.10). Area X was similar in location and dimension to that used by Tepper and Nurse (1997), Yackinous and Guinard (2002) and Fast et al., (2002). Area X also contains Area 1, which was reported in Part II as the best indicator of fungiform papillae counts in 7-10 year old children. Area Y was incorporated in this experiment because it contains the best predictor area (Area 5) for 11-12 year olds and adults in Part II. Thus, the 0.28 cm2 circle method was used in Part III to investigate whether Area X and/or Area Y would be suitable predictors of papillae density of the anterior tongue in different age groups.

3.14.1 Image Analysis

The images obtained in Part I were further analysed for this experiment. The 1 cm scale on the right hand side was used to calculate and draw a 0.28 cm2 circle on the tip of the tongue closest to the midline on the left side of the tongue (Area X). The second 0.28 cm2 circle was superimposed on Area Y (Figure 3.10). The fungiform papillae were then counted in both Areas and analysed.

105 Chapter 3

Part I Part II Part III

Y X

Figure 3.10: Images of the tongue showing the 2 x 1 cm bands (Part I), 8 small areas (Part II) and two 0.28 cm2 circles superimposed on an image showing Areas X and Y (Part III)

3.14.2 Statistical Analysis

For each subject group the number of fungiform papillae in the two 0.28 cm2 circles were counted (Appendix B, Tables B2- B5). A Pearson Correlation analysis (SPSS 12.0.1) was performed between the total fungiform papillae density in the 2 cm area from the tip of the tongue and papillae density in the two 0.28 cm2 circles in Areas X and Y to determine whether Area X and/or Area Y is a suitable predictor of papillae density on the anterior tongue (Appendix B, Table B15). In addition, two 2 X 4 (Sex X Group) between- participants ANOVAs were performed to determine if there were any significant differences between the fungiform papillae densities of the two 0.28 cm2 circles in Areas X and Y of all groups and between males and females of all groups (Appendix B, Tables B16 and B17).

106 Chapter 3

3.15 Results

3.15.1 Fungiform Papillae Density in Areas X and Y

The fungiform papillae density in Area X was 39.3  1.22 (S.E.), 39.0  1.62 (S.E.), 38.3  1.54 (S.E.) and 37.9  1.14 (S.E.) per 0.28 cm2 circle for 7-8, 9- 10, 11-12 year olds and adults, respectively. Lower densities were found in Area Y for all age groups and were 18.1  0.74 (S.E.), 17.7  0.77 (S.E.), 16.8  0.86 (S.E.) and 16.8  0.64 (S.E.) per 0.28 cm2 circle for 7-8 year olds to adults, respectively (Appendix B, Tables B2-B5).

3.15.2 Overall Analysis of Fungiform Papillae Density in Areas X and Y of All Age Groups

A Pearson Correlation analysis indicated there was a significant correlation between total fungiform papillae counts of the anterior tongue and Area X (R= 0.491: R2= 0.24 p<0.001) and Area Y (R= 0.325: R2= 0.11 p<0.001) when the data from all groups were combined (Appendix B, Table B15).

Additionally, the 2 X 4 ANOVA showed no significant difference in fungiform papillae density amongst age groups (F(3,107) = 0.251, p= 0.860) in Area X.

There were also no significant difference between males and females (F(1,107)

= 0.783, p= 0.378),and no significant Sex X Group interactions (F(7,107) = 0.398, p= 0.755) (Appendix B, Table B16). Similarly, for the data from Area Y the analysis showed no significant difference between age groups (F(3,107) = 0.594, p= 0.620) and no significant difference between males and females

(F(1,107) = 0.216, p=0.643). There were also no Sex X Group interactions

(F(7,107) = 0.077, p=0.972) (Appendix B, Table B17).

107 Chapter 3

3.15.3 Summary of Findings

The aim of Part III was to determine whether previous measures of papillae density using a 0.28 cm2 circle were good predictors of the total number of fungiform papillae in children and adults.

The aim was achieved and the analysis indicated a significant correlation between total fungiform papillae counts of the anterior tongue and both Areas X and Y when data for all subjects were combined. However both Areas X and Y correlations were lower than that for Area 1 (R=0.57: R2= 0.33, Table 3.6) when all groups were included in the analysis. Indicating the latter Areas produce acceptable less reliable densities than Areas 1 and 5 described in Part II.

108 Chapter 3

3.16 Discussion

The aims of Chapter 3 were (1) to determine at what age fungiform papillae density reaches adult levels (Part I), (2) to investigate whether fungiform papillae counts of small regions on the anterior tongue are suitable predictors of the fungiform papillae density of the whole anterior tongue (Part II), and (3) check whether previous methods correctly indicated the number of fungiform papillae on the anterior tongue (Part III).

The most important finding of Part I was that papillae density reaches adult levels by 9-10 years of age. Densities were highest in the 7-8 year olds, and decreased by 9-10 years of age where adult levels were reached. This is the first study that has demonstrated that fungiform papillae density, reaches adult numbers by mid-childhood. The findings of Part I are in accordance with previous reports (Moses et al., 1967; Segovia et al., 2002) that papillae density decreases from childhood to adulthood. Moses et al., (1967) showed that density decreased from 0-10 years olds to adults aged 50-60 years and Segovia et al., (2002) reported 8-9 year olds males had significantly higher papillae densities in two small areas on the tip of the tongue compared to adults. Interestingly, papillae reduction from childhood to adulthood has also been reported for circumvallate papillae (Arey et al., 1935). However no studies have indicated at what age circumvallate papillae density stabilizes.

The result of Part I showed that fungiform papillae numbers were significantly higher (13.6%) in 7-8 year olds than adults. Analysis of the papillae numbers in each of the two 1 cm regions indicated the difference was mainly due to the greater number of papillae in the most anterior region of 7-8-year old tongues (Table 3.1). There were also no changes in the fungiform papillae density in the first 1 cm in children above 7-8 years and no variations in the second 1 cm region with any age group, indicating that stabilization of fungiform papillae numbers in the anterior region of the tongue occurs by 9- 10 years of age. This is also in agreement with cessation of tongue growth in the anterior tongue (Temple et al., 2002).

109 Chapter 3

Another feature of the papillae that indicated changes in papillae density occurs until mid-childhood was their shape and size. In general, papillae of 7- 8 year olds were identified as small round structures and were larger in size in 9-10 year olds, whereas in 11-12 year olds and adults, they were irregular in shape. These results are also similar to those of Segovia et al., (2002) who reported using quantitative techniques that 8-9 year olds have rounder and smaller papillae than adults and adults have irregular shaped papillae. These changes in size and shape also suggest that development of the structure of papillae is complete at 11-12 years of age.

Interestingly Part I also showed that qualitative distribution patterns of fungiform papillae in children were different when compared to that of adults, particularly female adults and were classified into two types (Types A and B). Type A exhibited a decrease from childhood to adulthood for both genders, whilst Type B were more common in adults. The variation in distribution patterns found here could explain the finding by Stein et al., (1994) that children have higher sensitivity to tastants on small regions of the anterior tongue in comparison to adults.

Interestingly, psychophysical data indicate humans do not respond to the taste of salt similarly to adults until 2 years of age (Beauchamp et al., 1986) and at 8-9 years of age, male humans appear to have a poorer ability to perceive the constituents of taste mixtures (Oram et al., 2001). Therefore, even though 8-9 year olds have more papillae and more taste buds than adults these data suggest that neural maturation in not complete at this age.

Loss of fungiform papillae during childhood may also affect somatosensory perception on the tongue. Fungiform papillae receive both taste (chorda tymapani) and somatosensory (lingual nerve) afferents (Whitehead et al., 1985; Essicks et al., 2003). The gustatory afferents distribute into taste buds to synapse with taste cells (Nagy et al., 1982) and somatosensory afferents distributes around taste buds but terminate predominantly in the apex of papillae (Nagy et al., 1982; Whitehead et al., 1985). Such connections may

110 Chapter 3 therefore underlie the finding that humans with fewer papillae tend to have poorer spatial acuity (touch) than those who have significantly more papillae (Essick et al., 1999 and 2003).

Similarly, humans classified as super tasters have significantly higher number of fungiform papillae than medium and non-tasters (Bartoshuk et al., 1994; Mennella et al., 2005) and have a better lingual tactile acuity than other taster groups due to higher papillae numbers (Essicks et al., 2003). This could also suggest that children up to mid-childhood would expect to have superior spatial acuity than adults, indicating that children could also be more sensitive to chemical irritants like spicy/ hot tasting food and food granularity or texture, which affect food preference and selection. However further studies are required to explore the latter proposals.

The reduction and stabilization in papillae numbers from childhood to adulthood is not uncommon during the developmental stage in mammalian sensory systems. In olfaction, the initial formation of synapses and granule cells, the major interneurons in mouse olfactory bulb, is at 14 days of gestation. Thereafter the numbers of granule cells increases by 20 days of gestation followed by a large loss during the first 2 postnatal weeks, after which the number of granule cells and their synapse formation with mitral stabilizes (Hinds and Hinds 1976; Guthrie et al., 1990; Blanchart et al., 2006).

As to gustation the findings of this study showed that fungiform papillae on the anterior tongue decrease in number during mid-childhood and stabilize by 11-12 years of age. The decrease in papillae numbers would also include a decrease in the number of taste buds and cells within the papillae, which as previously reported (Segovia et al., 2002) 8-9 year olds have greater taste pores than adults. Other animal studies have also reported loss of papillae during developmental stages.

For example, studies on sheep reported that the number of fungiform papillae per receptor field increased from gestation to birth then decreased postnatally (Mistretta et al., 1988). It was also reported that the average

111 Chapter 3 number of taste bud per papillae increases then decreases during this period (Mistretta et al., 1988; Nagai et al., 1988). This increase and decrease during development in number of taste buds per receptive field and of taste buds per papilla coincide in time with an increase and then decrease in the number of chorda tympani nerve fibres (Nagai et al., 1988). Additionally in rhesus monkeys counts of optic nerve axons from early gestation to adulthood showed a rapid early prenatal increase of axons, followed by a rapid decrease during mid-gestation and a subsequent slower decrease and stabilization during the postnatal phase (Rakic and Riley, 1983).

As regards gender difference in the present studies, Part I showed there were no differences between the papillae densities of males and females in all groups, which was similar to the finding of Shahbake (2002) who reported no significant difference between 8-9 year old males and females. Other studies of adults have also reported no gender related differences in papillae and taste bud counts (Miller 1987; Zuniga et al., 1993).

Part II aimed at developing models for predicting papilla density on the anterior tongue and these are the only models currently available. In general the predictor models showed how well some small areas of the anterior tongue predict the overall papillae density of the anterior tongue. The main finding of Part II was that for children aged 7-10 years the best predictor of total papillae density was Area 1 at the tip of the tongue, and for 11-12 year olds and adults it was Area 5, an adjacent but more posterior region. A combination of fungiform papillae counts in Areas 1, 2 and 5 in 7-10 year olds children and Areas 5 and 1 in 11-12 year olds and adults produced modestly better predictor models. The finding that different Areas are better predictor models for younger age groups in comparison to those for 11-12 year olds and adults also indicates that stabilization of the distribution of papillae occurs by the age of 11-12 years.

Part III also showed that fungiform papillae counts in both Areas X and Y were higher in children than adults and importantly that the densities were similar to previous studies (Tepper and Nurse 1997; Fast et al., 2002;

112 Chapter 3

Yackinous and Guinard 2002). For example, calculation of papillae counts per cm2 indicated that the mean fungiform papillae densities of 7-8, 9-10, 11- 12 years old and adults were 139.05, 138.11, 135.49 and 133.95 papillae/ cm2 respectively in Area X, whilst Fast et al., (2002) and Yackinous and Guinard (2002) reported 99.5 and 141 papillae/ cm2 in adults. Similarly, Shahbake (2002) reported a mean value of 110 fungiform papillae/cm2 in 8-9 year children for an area similar in location and size to Area X.

Nevertheless, Areas X and Y produced less reliable densities than indicated by the R2 values for Areas 1 and 5 for children and adult groups found here. The lower predictability of Area X appears to be due to inclusion of part of Area 3 in the papillae counts, which was not one of the Areas found to be a predictor for any of the age groups in the 3 models described (Tables 3.2 and 3.3). Similarly, the R2 values for Area Y were lower than found for Area 5 for 11-12 year olds (0.52 vs. 0.28) and adults (0.46 vs 0.13) because part of Area 7 was included in the papillae counts and this area was not found to be a significant predictor for these age groups. Overall, Areas 1 and 5 were better predictors than Areas X and Y for particular age groups.

In conclusion, fungiform papillae on the anterior tongue decreased in number during mid-childhood and stabilized by 9-10 years of age. The sizes and the shapes of fungiform papillae of 11-12 year olds were also similar to adults. Additionally the differences in papillae numbers between adults and 7-8 year old children appears to be confined to the most anterior region of the tongue (Correa et al., 2013). Area 1 within this anterior region was the best predictor of overall number of papillae on the anterior tongue of children up to 10 years of age. On the other hand Area 5 was best predicator for 11-12 years olds and adults. Given that size, shape and distribution of fungiform papillae stabilizes at the age of 11-12 years old and that, adults and 8-9 year olds have similar number of taste pores per papilla (Segovia et al., 2002), it is suggested that that the peripheral gustatory system is fully functional by 11- 12 years of age (Correa et al., 2013).

113 Chapter 3

Finally, the results showed that measuring fungiform papillae density in a small area of the anterior tongue using the digital camera provides a rapid procedure for estimating the density of papillae on the whole anterior tongue (Correa et al., 2013). In the next Chapter, procedure involving the digital camera is modified for use in a clinical setting. In particular, a higher resolution digital camera is used to investigate whether it is possible to count fungiform papillae in humans without the use of a staining procedure, so as to provide a rapid truly non-invasive procedure, which minimizes discomfort to a patient.

114 Chapter 4

Chapter 4

Development of a Non-Staining Method for Counting Fungiform Papillae

115 Chapter 4

4.1 Introduction

Counts of fungiform papillae have been used in conjunction with psychophysical studies to explain the basis of individual variations in taste sensitivity or categorizing individuals based on their taste sensitivity to PROP (Miller and Reedy 1990a, b; Bartoshuk et al., 1994; Prutkin et al., 2000). The methods were used to quantify fungiform papillae are video-microscopy (Miller and Reedy 1990a, b; Segovia et al., 2002), contact endoscopy method (Just et al., 2006) and the digital camera method described in this thesis (Shahbake et al., 2005). However each of these techniques are invasive, since prior to capturing the images the chosen location is stained with either a food dye to facilitate fungiform papillae counting (Fast et al., 2002; Shahbake et al., 2005), or 0.5% methylene blue (Miller and Reddy 1990a, b; Segovia et al., 2002).

The staining procedure, however, is not always suitable for use in clinical situations where individuals may be suffering from an illness that causes damage to the tongue and staining is likely to be painful. For example, patients suffering from xerostomia (dry mouth syndrome) as a result of radiation therapy for head and neck cancer, or patients that have lesions on the tongue (Henkin et al., 1972; Mossman 1983; Sandow et al., 2006). Accordingly, there is a need for a method that does not involve staining the tongue of such patients.

To overcome the latter problem, the aim of Chapter 4 was to investigate whether fungiform papillae density can be quantified on the anterior tongue without the use of staining or drying procedures when a much higher resolution digital camera than that used in Chapters 2 and 3. This Chapter have been published in a peer reviewed journal; Pediatric Nephrology (Correa et al., 2015) (Appendix E, Paper 3).

116 Chapter 4

4.2 Aims

The aim of Chapter 4 was to determine if a non-staining procedure can be used in the evaluation of fungiform papillae density in humans in conjunction with a high resolution digital camera.

4.3 Methods

4.3.1 Participants

The subjects were 35 adults aged 20-24 years (26 females and 9 males; mean age = 21.63  0.27 S.E. years), who were students at the University of Western Sydney. The research was approved by the University Human Research Ethics Committee (Approval No. HEC02/188).

4.3.2 Equipment

The digital camera was a Canon AF SLR (EOS-1Ds) equipped with a CCD sensor, which produced digital images having a resolution of 12.0 mega- pixels (Figure 4.1). The camera was equipped with a macro lens (EF 100mm f/ 2.8 macro USM) and an extension tube (EF25). The extension tube was installed between the macro lens and the camera body and enabled close-up images at high magnification to be obtained by increasing the distance between the lens and the film plane (plane of focus). The film plane was the point in the camera where all the light rays converged, forming a sharp image. The configuration also included a macro ring light (MR-14EX) with two circular flash tubes, which consisted of an E-TTL auto-flash. The latter lighting arrangement was necessary because it enhanced the images taken at a close distance.

117 Chapter 4

Macro ring light with flash tubes

Figure 4.1: The Canon AF SLR (EOS-1Ds) camera with attachments

4.3.3 Experimental Procedure

The experimental procedure was conducted in three Stages and the tongue of each participant was photographed in each Stage. All three Stages were conducted on the same day to insure the same location was being photographed. The area photographed was equivalent to Area 5 that was described in Chapter 3, Part II, and as indicated earlier in Chapter 3 this location was chosen because it showed a high correlation with the total number of fungiform papillae in the anterior tongue of adults.

4.3.3.1 Stage 1; Tongue Wet and No-stain

At the commencement of the procedure, participants rinsed their mouth with deionized water (Milli-Ro-6 Plus System, conductivity 0.9 S). Each of them sat on a chair and supported their head by placing their arm on a table such that their chin protruded forward (Figure 2.4A). The subject then protruded their tongue and held it steady with their lips. Using a flexible ruler and a fine paintbrush containing red food dye (Roberts, RED 4R124 edible powder food colour), the most anterior border of Area 5 that was to be photographed (Figure 4.2A) was marked. A 10 X 3 mm wide piece of filter paper placed on the right side of the anterior tongue provided a scale to calculate the magnification of each image. The red mark remained in position for Stages 2

118 Chapter 4 and 3. Following this, 2-3 images of the anterior tongue, which included the marked area, were captured with the digital camera (Figure 4.2A). The 1 cm scale was removed from the tongue and the participant rested their tongue for 1 to 2 minutes.

4.3.3.2 Stage 2; Tongue Dried and No-stain

The participants were positioned as in Stage 1 but in this instance when they protruded their tongue it was dried with filter paper (Whatman’s No. 1). The 1 cm scale was replaced on the right anterior of the tongue and 2-3 images were captured (Figure 4.2B). The 1 cm scale was removed and the participant rested their tongue for 1 to 2 minutes.

4.3.3.3 Stage 3; Tongue Dried, Stained and Dried

The participants were positioned in the same manner as in Stage 1 and protruded their tongue, which was then dried as in Stage 2. A 6mm diameter filter paper (0.28 cm2 circle) was then dipped in 0.5% methylene blue (C152105), and using steel forceps the filter paper was applied above the red mark close to the midline and removed after 3 seconds (Figure 4.2C). The tongue was again dried with filter paper to prevent the methylene blue from spreading. As in the earlier stages the 1 cm scale was placed on the right side of the tongue. Following this, 2-3 images were captured from the stained area and analysed later.

119 Chapter 4

1 2

Marked with red food Filter paper scale added and the dye by a brush tongue photographed

(A) Stage 1

1 2

The tongue The tongue was photographed was dried with the scale in place (B) Stage 2

1 2

A 6mm diameter filter paper (0.28 cm2 circle) soaked in methylene The filter paper was removed blue was placed on the 1 cm line after 3 seconds close to the midline with forceps 3 4

The stained area was dried with The filter paper scale was placed a clean filter paper to prevent the on the right hand side and the dye from spreading tongue photographed (C) Stage 3 Figure 4.2: Diagrams of the procedures in the three Stages of tongue preparation for photography

120 Chapter 4

4.3.4 Image Analysis

The image analysis consisted of 2 procedures (4.3.4.1 and 4.3.4.2). All images were first assigned to 3 separate groups and numbered with 4 digit numbers, ensuring the identities of the participants were unknown to the experimenter until all the counts were completed and the data combined for statistical analyses.

4.3.4.1 Stage 1 and 2 Analyses

The images of the tongue were downloaded to a computer (Toshiba Satellite PRO6100) using the USB port and examined using an Adobe Photoshop 7.0 program. The midline of the tongue was drawn on the image using the “grid” line from the “view” menu. The 1 cm scale on each image was used to calculate and superimpose a 0.28 cm2 circle on the image where it was marked with the red food dye. The image was then magnified so that all the fungiform papillae were easily identifiable within the 0.28 cm2 circle (Figure 4.3A).

4.3.4.2 Stage 3 Analysis

After the images of the tongue were downloaded into the computer, the midline of the image was marked as in the latter analyses. A 0.28 cm2 circle was superimposed on the stained area as in some cases the dye had spread outside the stained area. The images were magnified and the fungiform papillae in the stained area were counted and recorded for statistical analysis (Figure 4.3B).

121 Chapter 4

(A)

(B)

Figure 4.3: The image preparation before the fungiform papillae count in (A) Stages 1 and 2 (B) Stage 3

4.3.5 Statistical Analysis

The total number of fungiform papillae in the 0.28 cm2 circle was counted for each participant for each of the three Stages (Appendix C, Table C1). A Bland–Altman (B-A) plot was then used to establish the 95 % Limits of Agreement (LoA) between the papillae densities for each pair of Stages; 1 versus 2, 1 versus 3, and 2 versus 3 (Appendix C, Tables C2 to C4) and three regression analyses determined whether the limits of agreement were different between the three stages (Appendix C, Table C5).

The B-A plot is used when a conventional standard method is compared with new method, so it is a comparison between two methods with reported LoA. B-A plot measures systematic difference between two methods and also

122 Chapter 4 identifies possible outliers. If there are multiple methods to compare, like comparing the 3 Stages of staining then 3 sets of analyses are required; (1) Stage 1 versus 2, (2) 1 versus 3 (3) 2 versus 3.

4.4 Results

4.4.1 Description of Fungiform Papillae in Each of the Three Stages

In Stage 1 (wet and no-stain) the fungiform papillae were elevated bright pink and red structures, which were larger than the surrounding filiform papillae and easily distinguished from the background surface of the tongue (Figure 4.4A). In some of the fungiform papillae clear images of blood vessels were observed (Figure 4.5). Interestingly, many distinctive types of blood vessels were observed, some had dotted patterns (Figure 4.5A) and some were branched (Figure 4.5 B, C, D).

In Stage 2 (dried and no-stain) the fungiform papillae were pale pink in colour and some of the smaller fungiform papillae were flattened in comparison to their appearance in Stage 1. In some instances it was difficult to identify the smaller fungiform papillae because they blended in with the surrounding filiform papillae (Figure 4.4B, D and E). However, even though the papillae were sometimes more difficult to identify than when wet, it was possible to identify them.

In Stage 3 (dried, stained and dried) the fungiform papillae were elevated similar to Stage 1, but not all the papillae were stained with methylene blue. In some participants stained fungiform papillae were darker blue (Figure 4.4F) whereas, in others the fungiform papillae were slightly stained (Figure 4.4G). However, in all images the surrounding tissues were always darker in colour than the fungiform papillae. In Stage 3 fungiform papillae identification was simpler than in Stage 2 because the surrounding tissue and the filiform papillae were stained darker.

123 Chapter 4

(A) (B) (C)

(D) (E)

(F) (G)

Figure 4.4: Fungiform papillae of participant in (A) Stage 1, (B) Stage 2 and (C) Stage 3. In (D) and (E) arrows show hidden fungiform papillae in Stage 2. (F) Dark stained papillae and surrounding tissue and (G) light stained papillae and surrounding tissue in Stage 3

124 Chapter 4

(A)

(B) Figure 4.5: Types of blood vessel patterns of fungiform papillae. (A) Dotted pattern and (B) branched pattern. Arrows indicate blood vessels

125 Chapter 4

(C)

(D)

Figure 4.5 (cont.): (C) Branched pattern and (D) branched and lopped pattern. Arrows indicate blood vessels

126 Chapter 4

4.4.2 Counts of Fungiform Papillae in the Common 0.28 cm2 Circle of the Three Stages

The mean number of fungiform papillae in the 0.28 cm2 circle for images from Stages 1, 2 and 3, respectively, were 10.94  0.57 (S.E.), 11.00  0.59 (S.E.) and 11.14  0.66 (S.E.) (Appendix C, Table C1).

The 95% LoA (-3.06, 2.94) contain 94% (33/ 35) of the difference scores of fungiform papillae counts for Stage 1 and 2 (Figure 4.6), and 97% (34/ 35) for both Stage 1 and 3 (LoA: -4.94, 2.54) (Figure 4.7) and Stage 2 and 3 (LoA: - 4.16, 1.87) (Figure 4.8). The mean difference of the measurements in three sets of analysis were -0.057 ± 0.26 (S.E.), -1.20 ± 0.32 (S.E.) and -1.14 ± 0.25 (S.E.) for Stage 1 versus 2, 1 versus 3 and 2 versus 3 respectively (Appendix C, Tables C2 to C4).

In addition 3 linear regression analyses were also conducted for all three B-A plots to test whether there were any trends, and also to investigate where there might be more data points below or above the mean difference line (Figures 4.6 to 4.8). In all three analysis there were no significant (p= 0.766, p=0.108 and p=0.094 for Stage 1 versus 2, 1 versus 3 and 2 versus 3) difference between the Stages indicating there seems to be a level of agreement between the 3 Stages and any of the three methods could be used to count fungiform papillae. (Appendix C, Table C5).

127 Chapter 4

Figure 4.6: B-A difference plot of fungiform papillae counts between Stage 1 and 2

Figure 4.7: B-A difference plot of fungiform papillae counts between Stage 1 and 3

128 Chapter 4

Figure 4.8: B-A difference plot of fungiform papillae counts between Stage 2 and 3

4.4.3 Summary of Findings

The statistical analyses indicated that similar numbers of fungiform papillae were identified in all three Stages. Of importance was the finding that it was possible to identify the fungiform papillae from the surrounding filiform papillae in Stage 1 (no stain and wet) and Stage 2 (no stain and dried), even though in Stage 2 some of the smaller fungiform papillae were more difficult to identify because they had been flattened during the drying process (Figure 4.4D and E). In Stage 3 (stained and dried) it was simpler to identify fungiform papillae from the surrounding filiform papillae than from Stage 2 because the fungiform papillae were stained. Overall, the results indicated that it was possible to quantify fungiform papillae density on the anterior tongue without using a staining or drying procedure, thereby achieving the aim of Chapter 4 (Correa et al., 2015).

129 Chapter 4

4.5 Discussion

The aim of Chapter 4 was to investigate, whether it is possible to quantify fungiform papillae density on the anterior human tongue without using a staining procedure. The goal was important, since success had the potential to open up opportunities for conducting studies on monitoring taste loss and recovery in many illnesses that is currently not possible with other methods. The results of Chapter 4 showed that there were no significant differences between the mean number of fungiform papillae counted in the three stages using the high-resolution digital camera. Previously (Miller and Reddy 1990 a, b) the tongue was dried and stained before the images were captured. When this was not suitable for participants, either the study was discontinued or the participants were not included in the study. Thus, development of the new method for counting fungiform papillae without staining the tongue overcomes this short-coming and allows the use of anatomical changes on the tongue to provide an indication of potential taste loss.

It should be noted that initially red food dye was used as a marker for the 0.28 cm2 circle location during the development of the non-staining method. Now that the method has been shown to work the food dye is not necessary, but, it is important to use the 1 cm scale for calculation of the 0.28 cm2 circle that is superimposed on captured images.

Furthermore, in circumstances where a participant cannot tolerate the 1 cm paper scale being placed on their tongue (for example patients with tongue ulceration or xerostomia) and/ or the very small amount of red marker dye, an alternative is to use the digital grid method (Chapter 3, Correa et al., 2013). The latter method can be used to identify the location and size of the 0.28 cm2 circle on the anterior tongue with a mean error rate of 5.94% ± 0.46 (S.E.) in comparison to using the same image containing the 1 cm scale.

The critical finding of this experiment was that it was easier to distinguish fungiform papillae when the tongue surface was not dried, as the papillae

130 Chapter 4 were more elevated. When the tongue was dried some of the fungiform papillae were flattened (Figure 4.4D and E), making it more difficult to identify some of the papillae. The images captured in Chapter 4 also showed various types of blood vessels in the fungiform papillae. These appeared to be similar to the report by Negoro et al., (2004) who classified the blood vessels of the fungiform papillae into five groups (Figure 1.15). In the present study 3 of the 5 types were observed; Type A was classified as having a clear loop and wooden branch shape (Figure 4.5B), Type B; unclear loop and wooden branch shape (Figure 4.5A and C) and Type C; elongated blood vessels (Figure 4.5D and E). The other two types were not observed here.

In summary, fungiform papillae were identified on the surface of the tongue using a high-resolution digital camera without staining or drying the tongue. It was also possible to observe blood vessels of fungiform papillae in unstained and wet images of tongue captured. The satisfactory outcome of the present study provides a stepping-stone to develop other methods to monitor taste loss and recovery in patients suffering from various illness or medical interventions, where it may not be possible to stain the tongue surface (Correa et al., 2015). Chapter 5 demonstrates that this method can be used to monitor taste loss in patients suffering from chronic renal disease.

131 Chapter 5

Chapter 5

The Effect of Chronic Renal Failure on Fungiform Papillae Density in Children

132 Chapter 5

5.1 Introduction

5.1.1 Renal Failure

Renal failure is the inability of the kidney to excrete wastes, concentrate urine and conserve electrolytes (Bohatyrewicz et al., 2005; Martini 2006). Renal failure may be acute or chronic. Acute renal failure (azotemia or uremia) is characterized by oliguria and by the rapid accumulation of nitrogenous wastes in the blood (Laville and Fouque 2000). It results from haemorrhage, trauma, burn, and toxic injury to the kidney, acute pyelonephritis or lower urinary tract obstruction (Agarwal et al., 2003; Moreau and Lebrec 2007). Most acute renal failures are reversible after the underlying cause has been identified (Trof et al., 2006; Oh et al., 2006). Acute renal failure may have three typical phases: prodromal, oliguric and post-oliguric (Ko 1992; Coe and Lail 2007). Treatment includes restricted intake of fluids and of all substances that require excretion by the kidney (Martini 2006).

Chronic renal failure (CRF) may result from many other diseases like polycystic kidney disease (Douma and Smit 2006; Sarafidis et al., 2008). The early signs include sluggishness, fatigue and mental dullness (Cassileth et al., 1984; Miyata et al., 1997; De Santo et al., 2006). Later anuria, convulsions, gastro-intestinal bleeding, malnutrition and various neuropathies may occur and the skin may turn yellow-brown (Miyata et al., 1997; Douma and Smit 2006). An increase in the amounts of urea and creatinine occur and a constant volume of urine is voided regardless of variations of water intake (Pljesa et al., 2005; Douma and Smit 2006). Treatment usually includes restricted water and protein intake and the use of diuretics (Coleman et al., 2002; Huang et al., 2008). When medical measures have been exhausted, long-term dialysis is often begun, and kidney transplantation is considered (Martini 2006).

Dialysis is a medical procedure for the removal of certain elements from the blood or lymph by virtue of the difference in their rates of diffusion through an

133 Chapter 5 external semipermeable membrane (Douma and Smit 2006). Dialysis may be used to remove poisons and excessive amounts of drugs, to correct serious electrolytes and acid-base imbalances, and to remove urea, uric acid and creatinine in cases of chronic end-stage of renal disease (Douma and Smit 2006; Teixeira et al., 2008). Dialysis involves diffusion of particles from an area of high to lower concentration, osmosis of fluid across the membrane from an area of lesser to one of greater concentration of particles and ultrafiltration, or movement of fluid across the membrane as result of an artificially created pressure differential (Douma and Smit 2006). There are two types of dialysis; hemodialysis and peritoneal dialysis (Douma and Smit 2006; Martini 2006).

Hemodialysis is a procedure in which impurities or wastes are removed from the blood (Kho et al., 1999; Laville and Fouque 2000; Leshem et al., 2003). It is used in treating patients with renal failure and various toxic conditions (Laville and Fouque 2000). The patient’s blood is shunted from the body through a machine for diffusion and ultrafiltration and then returned to the patient’s circulation (Douma and Smit 2006). Hemodialysis requires access to the patient’s blood stream and a mechanism for the transport of the blood to and from the dialyzer machine to the dialyzer patient (Figure 5.1) (Douma and Smit 2006).

Figure 5.1: Diagram of hemodialysis set up (Martini 2006)

134 Chapter 5

Peritoneal dialysis is a dialysis procedure performed to correct an imbalance of fluid or of electrolytes in the blood or to remove toxins, drugs, or other wastes normally excreted by the kidney (Middleton et al., 1999; Douma and Smit 2006). The peritoneum is used as a diffusible membrane. Peritoneum is the natural lining of the abdomen, which, acts as the dialysis membrane (Martini 2006). Peritoneal dialysis may be performed nightly for chronically ill children while they sleep and also may be carried out regularly at home (Figure 5.2) (Hurley et al., 1987; Dobell et al., 1993).

Figure 5.2: Diagram of peritoneal dialysis set up (Martini 2006)

5.1.2 Effect of Renal Failure on Taste Sensitivity

A number of studies have reported that people suffering from renal failure have impaired taste acuity (Fernstrom et al., 1996; Middleton et al., 1999; Matson et al., 2003). Patients with renal failure may not consume an adequate dietary intake for many reasons; changes in taste perception (Matson et al., 2003), restricted diets, nausea, vomiting, anaemia (Middleton et al., 1999) and depression (Cassileth et al., 1984; Miyata et al., 1997). Patients frequently complain about loss of appetite and the dull taste of food, and often report a persistent and unpleasant taste sensation (Ciechanover et al., 1980). In several studies patients with renal disease have demonstrated

135 Chapter 5 raised thresholds for some or all of the four primary tastes (sweet, sour, salt and bitter) (Atkin-Thor et al., 1978; Burg et al., 1979). One of the reasons for taste loss that has been reported is zinc deficiency (Matson et al., 2003).

In general, renal failure patients have a poor appetite. Anorexia, hypogeusia and poor growth are major features of zinc deficiency (Cabral et al., 2005). Levels of zinc (Zn) in the hair have been reported to provide a good index of stores of Zn in the body (Hambidge et al., 1972; Cabral et al., 2005; Dvornik et al., 2006). In addition to the level of Zn in hair others have measured the amount of Zn in plasma, serum and red cells in blood (Henkin et al, 1971; Cabral et al., 2005). Abnormal Zn levels have been reported in dialysis patients. Atkin-thor et al., (1978) studied 20 adult chronic hemodialysis patients and 9 controls (healthy volunteers), all dialysis patients suffered varying degrees of nausea and anorexia. The control group was given 220 mg of Zn and the patients were given 440 mg Zn. Hair and serum Zn concentrations were measured before and after Zn supplementation. Before supplementation 95% of dialysis patients tested had some degree of hypogeusia. Significant improvement in taste acuity occurred after 6 weeks. Similar findings in increased taste acuity by increasing Zn levels were reported in later studies (Blendis et al., 1981; Mafra et al., 2002; Dvornik et al., 2006).

5.1.3 Effects of Renal Failure on Taste Bud Density

Many factors can affect taste sensitivity, one anatomical feature which was affected in patients with renal disease were peripheral taste buds. Astback et al., (1999) studied the number of taste buds and the neural patterns in taste buds with the use of specific markers for peripheral nervous tissue. The gustatory neural patterns (gustatory neural circuits) are generated by a neuromatrix, they are spatially organized across a neural population to represent information about various taste qualities (for example, “sweet”, “salty”, etc.) (Lemon and Katz 2007). The study included 36 (9 females and 27 males, mean age = 56 years) adults with CRF (17 patients had not started dialysis, 12 patients were on peritoneal dialysis and 7 patients on

136 Chapter 5 hemodialysis) 19 renal transplant recipients and 40 healthy adult subjects (11 males and 29 females, mean age = 30 years) which were not matched for gender or age. From each subject 2 or 3 fungiform papillae were dissected from the anterior tongue and sectioned under a light microscope to determine whether they contained taste buds. The patients with CRF had significantly fewer taste buds than the controls. For example, 78%, 79% and 12% of the fungiform papillae examined in CRF patients, renal transplant recipients and the control group did not contain taste buds. Thus, it was concluded that lack of taste buds might account for the impaired taste acuity in subjects with CRF, but the cause of less taste buds in patients was not explained.

Currently the study by Astback et al., (1999) is the only one that has investigated the anatomy of the taste system in patients undergoing dialysis or suffering from renal failure. However, the latter study did not report structural differences in the taste papillae or taste buds between the control group and the patients, only that there were less taste buds. Importantly, for the purpose of the present study the use of this invasive method to count the papillae precludes its use with children. This Chapter have been published in a peer reviewed journal; Pediatric Nephrology (Correa et al., 2015) (Appendix E, Paper 3).

5.2 Aims

The aims of Chapter 5 were to (1) investigate the effect of CRF on fungiform papillae density of children with the use of the high-resolution digital camera and non-staining procedure, and (2) whether there were any structural features of the papillae that differentiate the CRF patients from the clinical controls.

137 Chapter 5

5.3 Methods

5.3.1 Participants

All 24 participants (6 females and 18 males aged 5 to 17 years, median = 12 years) were patients of the Department of Nephrology, Sydney Children’s Hospital who regularly visited the Outpatient’s Clinic. The patient’s specialist obtained the consent of the parents/ guardians of the children prior to the commencement of the study. In addition, all parents were present for the duration of the experimental procedure and provided written consent for the children to participate. The children were also asked if they wished to participate. Approval for the study was obtained from the South Eastern Sydney and Illawarra Area Health Service Human Research Ethics Committee (Approval Number HREC 05/007).

5.3.1.1 Classification of the Patients

CRF is defined by two criteria; (1) kidney damage for  3 months, as defined by structural or functional abnormalities of the kidney with or without decreased estimated Glomerular Filtration Rate (eGFR) and can be manifested either by; pathological markers of kidney damage, including abnormalities in the composition of the blood or urine, or abnormalities in imaging tests. (2) eGFR < 60 mL/min/1.73 m2 for  3 months with or without kidney damage (British Renal Foundation 2006). eGFR is the total rate of filtration of blood by the kidney. It is calculated by a number of formulas based on the patient’s gender, race, age and health conditions.

The equation used to calculate the eGFR in the present study was based on Schwartz and Furth (2007) and is currently used in the Department of Nephrology, Sydney Children’s Hospital. The patients were divided into two groups based on the eGFR value. One group had a eGFR<60 and the other eGFR>89. The classification was based on the North America Kidney Disease Outcome Quality Intuitive (K/DOQI) (2002) guidelines, Caring

138 Chapter 5

Australia with Renal Impairment guidelines (CARI) (2005) and the British Renal Foundation (BRF) (2006) (Table 5.1).

Group eGFR Description Number of (mL/min/1.73m2) participants 1: Clinical Controls eGFR > 89 Kidney damage with normal 12 to high eGFR 2: Chronic < 60 Moderate kidney impairment 12 with low eGFR

Table 5.1: Description of patient and clinical control groups

Based on the different stages of kidney failure, the clinical controls were considered as having kidney damage but not kidney failure, whereas, the patient group had CRF. Clinical controls were included Instead of healthy controls because of their availability and also previous study by Armstrong et al., (2010) have reported no significant differences between taste function in healthy and clinical controls. In addition no previous studies of children or adults with various types of renal disease or early stage of CRF have been found to demonstrate improved taste function. Hence, it was assumed that inclusion of patients with renal disease in the control group would not reduce the chance of finding a difference in taste function when compared to CRF patients.

The patient and the clinical control group were age and gender matched with two exemptions for gender (Table 5.2). The researcher who photographed the tongues was unaware of the children’s eGFR and underlying medical condition at the time of the study.

139 Chapter 5

Patient Age M/F# eGFR Diagnosis Control Age M/F# eGFR Diagnosis

1** 14 M 56.7 PUV* 1 15 F 161 MCGN*

2 14 M 58 Dysplastic 2 15 M 141 HSP*

3 16 M 20 RN* 3 16 M 109 Hypertension

4 12 M 34 PUV 4 11 M 94 Nephrotic

5 6 M 32 Dysplastic 5 7 M 127 Hypertension

6 16 M 25 RN 6 18 M 104 RPGN*

7 6 M 47 Dysplastic 7 8 M 97 Post BMTN*

8 17 M 11.6 Dysplastic 8 17 M 101 FSGS*

9 11 F 15 Dysplastic 9 11 F 112 UTI*

10 15 M 9.6 Dysplastic 10 15 M 174 FSGS

11 9 F 14 Dysplastic 11 8 F 125 RN

12** 6 M 23 Dysplastic 12 5 F 113 RN

Table 5.2: Clinical characteristics of age and gender matched patients and controls # M= male and F= female *BMTN: Bone Marrow Transplant Nephropathy HSP: Henoch-Schonlein Purpura PUV: Posterior Urethral Valves FSGS: Focal Segmental Glomerulosclerosis MCGN: Mesangiocapillary Glomerulonephritis RN: Reflux Nephropathy RPGN: Rapidly Progressive Glomerulonephritis UTI: Urinary Tract Infection **Gender difference

5.3.2 Experimental Procedure

A 1 cm scale was placed on the right anterior tongue and the images of the anterior tongue captured (Figure 5.3(1)). The selected area was defined by using the 1 cm scale to measure from the tip of the tongue to the desired area. The area was subdivided into 8 small areas as per Chapter 3 Part II (Figure 5.3(2)). The Area chosen to count fungiform papillae was Area 5. This area was chosen because 16 out of 24 participants were aged over 10 years and according to the finding of Chapter 3, Part II, papillae density of Area 5 was the best predictor for participants over 10 years old and

140 Chapter 5 according to Chapter 3, Part III it also produced a significant correlation with the total number of fungiform papillae in all children and adults. The 0.28 cm2 circle was drawn based on the 1 cm scale and superimposed on the captured images (Figure 5.3(3)).

141 Chapter 5

(1) A 1 cm scale was placed on the right anterior tongue and images were captured.

(2) The outline drawing of the 1 cm scale is superimposed on the

image and the tongue image subdivided into 8 small areas (Chapter 3, Part II)

(3) The 1 cm scale is used to calculate and draw a 0.28 cm2 circle superimposed on the image

Figure 5.3: Diagram of the procedure for superimposing the 0.28 cm2 circle on the tongue using the 1 cm scale

142 Chapter 5

5.3.3 Statistical Analysis

For each subject the fungiform papillae were counted and recorded in the superimposed 0.28 cm2 circle (Appendix D, Table D1). Two Kruskal-Wallis tests followed by post hoc Mann-Whitney tests were conducted to investigate whether there was a significant difference in fungiform papillae density and eGFR between the patient and control groups (Appendix D, Tables D2 to D5). Additionally, a Pearson correlation (2-tailed) analysis was conducted to investigate whether there was a relationship between the eGFR and the fungiform papillae density of the patient and control groups (Appendix D, Table D6).

5.4 Results

5.4.1 Statistical Analysis of Fungiform Papillae Density and eGFR in Control and Patient Groups

The fungiform papillae in the 0.28 cm2 circle were counted as described in Chapter 4. The median number of fungiform papillae for the control group was 31.84 with a range of 7.07 to 53.06 papillae, whilst the patient group median was 19.46 with a range of 14.15 to 31.84 papillae/ cm2 (Appendix D, Table D1). The median and range of eGFR for the control and patient groups, respectively, were 112.5, 94 to 174 and 24, 9.6 to 58 mL/min/1.73 m2. Two Kruskal-Wallis tests showed there were significantly less fungiform papillae density (X2=7.17, p = 0.007) and eGFR (X2= 17.28, p < 0.001) in the patient group than in the control group (Appendix D, Tables D2 and D3). The post hoc Mann–Whitney tests also confirmed that there was a significant difference between the control group and the patients for papillae density (U=26.0, p=0.0068) and eGFR (U=0.00, p=0.000). (Appendix D, Tables D4 and D5). Furthermore, there was a significant correlation between the fungiform papillae density and eGFR (R= 0.56, p = 0.004), with the number of papillae decreasing significantly with lower eGFR (Figure 5.4) (Appendix D, Table D6).

143 Chapter 5

 Control group  Patient group

Figure 5.4: Relationship between eGFR and fungiform papillae density of patient and control groups

144 Chapter 5

5.4.2 Topographical Variation of Fungiform Papillae in Control and Patient Groups

Visual examination of the papillae in the captured images of both participant groups revealed a number of features. In some patients (2/ 12), the surface of the tongue was covered with yellow fur (Figure 5.5). Figure 5.6 shows a typical image of patient and control tongues. In all images the area that the fungiform papillae were counted was marked with a dotted circle. The patient and control groups were compared for colour of the tongue surface, fur coating on the tongue and fungiform papillae visibility (Table 5.3). In the patient group, the surface of the tongue was generally pale pink in colour and dry. Furthermore, the fungiform papillae were very difficult to view at low magnifications. In contrast the tongue surfaces of the control group were mainly bright pink and did not appear dry, and there was no fur coating on tongue surfaces of any of the control group. Additionally the fungiform papillae appeared elevated and were easily identified from the surrounding filiform papillae (Figure 5.6).

145 Chapter 5

Figure 5.5: Tongue of patient 10. White arrow shows yellow fur coating on tongue surface

146 Chapter 5

Patient Control

Figure 5.6: Typical tongues of (A) patient and (B) control. The white dotted circle indicates the area within which fungiform papillae density was measured. Arrows show fungiform papillae. Scale = 5mm.

147 Chapter 5

A summary of the data of topographical tongue characteristics of patients and their matched controls is given in Table 5.3. The table shows that, 67% (8 out of 12) of patients could be described as having a pale pink tongue surface in comparison to 17% (2 out of 12) of the controls. In 58.3% of the patients the fungiform papillae were less visible, in particular they were flatter. Furthermore, 17% (2 out of 12) of the patients had a yellow fur coating on the tongue surface. On the other hand, one patient (Figure 5.7; patient 12) diagnosed with end stage CRF had fungiform papillae that appeared different in comparison to others in the group, in that there was a large cluster of fungiform papillae on the anterior tongue and the fungiform papillae were irregular in shape (Figure 5.7).

Fungiform papillae Colour (tongue invisibility & Flat Fur Coating Pair # surface) shape Control Patient Control Patient Control Patient 1 P* PP* - - - - 2 P P - Yes - - 3 P PP - - - - 4 P P - - - 5 P PP - Yes - - 6 PP PP - Yes - - 7 P PP - - - - 8 P PP - Yes - - 9 P PP - Yes - - 10 P PP - Yes - Yes 11 P P - Yes - Yes 12 PP P - - Total PP (/12) 2 8 - Total P (/12) 10 4 Total FP* 0 7 invisibility (/12) Total Fur (/12) 0 2

Table 5.3: Tongue topographical characteristics of age and gender matched patients and controls *PP= pale pink, P= pink, FP= Fungiform papillae

148 Chapter 5

Figure 5.7: Tongue of patient 12. White arrows show examples of enlarged and irregular fungiform papillae

149 Chapter 5

5.4.3 Summary of Findings

In summary the most important finding was that the fungiform papillae density in patients were significantly lower than in controls and the counts decreased with decreasing eGFR values (Correa et al., 2015). The results also showed that there were topographical differences on the tongue surface between the control and patient groups. Most of the patients exhibited pale pink tongue surfaces when compared to their matched controls. In the control group papillae were readily visible and easy to identify whereas in 7 out of 12 patients the fungiform papillae were harder to observe and appeared flatter.

150 Chapter 5

5.5 Discussion

The use of the new non-invasive method was demonstrated in Chapter 5, where the effects of renal disease on the morphology and the density of fungiform papillae on the anterior tongue of children were investigated. The location where the fungiform papillae were examined corresponded to the location described in Chapter 4. The critical finding was that the number of papillae was significantly lower in the patients than the controls, which paralleled the study by Astback et al., (1999), who found that 78% of the fungiform papillae of patients did not contain taste buds in comparison to only 12% of the papillae of the control group. The present and past data, therefore, indicate that patients with renal disease have lower numbers of papillae or taste buds, which would be expected to decrease their ability to perceive tastes. Furthermore, there was a significant correlation between the fungiform papillae density and eGFR, which is in agreement with a an earlier psychophysical study by Armstrong et al., (2010) who reported a strong correlation between taste loss as measured by the ability to identify tastes and eGFR, with decreasing values of eGFR correlating with a poorer ability to identify sweet, bitter and sour tastes.

In the present study in addition to the decrease in the number of fungiform papillae in CRF patients, there were other variations in the morphology of the papillae. In the patient group the tongue surface was usually pale pink (Figure 5.6) in comparison to the controls, which were pink. In patient number 12 the fungiform papillae were very visible and irregular in shape in comparison to the matched control (Figure 5.7). This irregular shape and large papillae were only seen in patient 12 and was not seen any other matched pairs. In the previous study by Negoro et al., (2004), the number of fungiform papillae with abundant vessels tended to be less on the tongue of patients with taste disorders than on normal tongues. In that study fungiform papillae were classified into 4 shape types (Figure 1.15). It was described that atrophic papillae have a flat surface (Type 4) and the blood vessels appeared either as having a granular shape (Type D) or unclear and absent

151 Chapter 5 blood vessels (Type E) (Figure 1.15). Type 4 was similar to those found in Chapter 5 where in 58.3% of the patients the fungiform papillae appeared flatter in comparison to their matched control. Fungiform papillae were also less visible and harder to identify in patients than their matched controls. Conceivably, the decrease or disappearance of blood vessels and less visible fungiform papillae could be two of the signs of degenerating taste buds, which may be affected centrally and/ or peripherally. For example, a reason for the loss of fungiform papillae could be due to loss of neurotrophic support, which is important in development of fungiform papillae and their taste buds (Ganchrow et al., 2003a; Oakley and Witt 2004). There are four main mammalian neurotrophins: (1) nerve growth factor (NGF), (2) brain derived neurotrophic factor (BDNF), (3) neurotrophin-3 (NT3), and (4) neurotrophin-4 (NT4) (Oakley and Witt 2004). In animal studies it has been reported that BDNF and its receptor tyrosine kinase B (TrkB) are essential molecular factors for normal nerve development in the gustatory system and the absence or over-expression of BDNF or TrkB may cause a decrease in taste bud density, which leads to atrophic papillae (Ganchrow et al., 2003b).

In conclusion, as demonstrated in Chapter 5 the non-staining method can be used to monitor taste loss in patients suffering with CRF indicating that it can be used with patients with other diseases suspected of affecting taste. Importantly, the method showed that CRF patients have significantly lower numbers of fungiform papillae than controls and that there were many variations between the appearance of fungiform papillae of patients and their matched controls. In addition, there was a significant correlation between the eGFR and fungiform papillae counts suggesting that there may be a critical value of eGFR that indicates taste loss is occurring (Correa et al., 2015).

152 Chapter 6

Chapter 6

General Discussion

153 Chapter 6 General Discussion

The two primary objectives of this thesis were to design a portable and more efficient method than currently available to quantify fungiform papillae on the human tongue (Shahbake et al., 2005) and to investigate fungiform papillae density during development in humans (Correa et al., 2013).

As regards the first objective, this was successfully achieved using both low (Shahbake et al., 2005) and high-resolution digital cameras to capture the images of fungiform papillae (Correa et al., 2015). The digital camera based method was developed for use outside and inside the laboratory such as with humans of different ages in developmental studies, and in hospital at a patient’s bedside or outpatient clinics. The method was also used to count fungiform papillae via non-staining and non-drying histological procedures (Correa et al., 2015).

Since the development of the digital camera method in this laboratory others have used the procedure for counting fungiform papillae in adults (Wockel et al., 2008; Zhang et al., 2009). For example, Wockel et al., (2008) studied fungiform papillae numbers in patients (14-20 years of age) with eating disorders and reported that patients with a disorder had significantly less fungiform papillae in comparison to their matched controls. Additionally, the relationship between fungiform papillae density and threshold of sucrose was investigated in male adults (18-23 years old) by Zhang et al., (2009), who reported a negative correlation between sucrose thresholds and fungiform papillae density.

In the present thesis in Chapter 2 validation of the digital camera method was achieved (Shahbake et al., 2005) by comparing it with data obtained using video-microscopy, the method that had been commonly used for quantifying fungiform papillae and taste buds on the human tongue prior to the development of the digital camera method. The video-microscope has been used in many studies for papillae quantification since 1989 (Miller and Reedy

154 Chapter 6

1990a, b; Segovia et al., 2002; Just et al., 2006), however, it is restricted to use in a laboratory and is a bulky and time-consuming procedure. The results of Chapter 2 showed there was no significant difference between counts of fungiform papillae using the two methods and demonstrated the potential of the digital camera as a rapid and portable method (Shahbake et al., 2005). The main advantages of using a digital camera, are that in addition to providing an efficient, portable and cost effective method, it provides a reduction of 60% in the time required for equipment configuration to compliment image analysis, a very significant time-saving in studies with large groups of participants. The potential to use large groups of subjects in many studies will also assist in increasing the statistical power of studies. Furthermore, the portability of the camera and ease of use should see it utilized in taste perception studies that are conducted in parallel with taste anatomy and physiology measures.

In Chapter 3 (Part I) the digital camera method was further utilised to quantify fungiform papillae density during human development and investigate when the number of papillae reaches adult levels. This study was the first to demonstrate that the fungiform papillae density on the human tongue reaches adult levels by mid-childhood (Correa et al., 2013). Overall the results showed a significant difference of papillae counts between 7-8 year olds and adults and that similar counts to those recorded for adults were achieved in 11-12 year olds. The results were in accordance with the previous reports of Segovia et al., (2002) and James et al., (1997). Segovia et al., (2002), used videomicroscopy to show that 8-9 year old males had a significantly higher papillae density in two small areas on the tip of the tongue in comparison to adults. On the other hand James et al., (1997) showed 8-9 year old male children had significantly poorer ability to detect sucrose, sodium chloride and citric acid than female adults which indicates gustatory systems of males is not fully mature for detection of tastants at 8-9 years of age.

The results of Chapter 3 (Part I) showed stabilization of papillae numbers occurred by 9–10 years of age, which coincides with completion of the

155 Chapter 6 growth of the “taste-sensitive” area of the anterior tongue (Temple et al., 2002). The finding that there were no changes in the numbers of papillae in the first 1 cm interval in children aged above 7–8 years and no changes in the second 1 cm interval with any of the age groups indicates that stabilization of papillae numbers in the anterior tongue had occurred by 9–10 years of age.

Another feature of the papillae that indicated changes in papillae density occurs until mid-childhood was their shape and size. Papillae of 7–8 year olds were identified as small round structures; in 9–10 year olds, they were larger in size, whereas in 11–12 year olds and adults, they were irregular in shape. These results are in agreement with the report that 8–9 year olds have rounder and smaller papillae than adults, and adults have papillae that differ in shape more than in children (Segovia et al., 2002). The changes in size and shape also suggest that development of the structure of papillae is complete at 11–12 years of age.

Chapter 3 (Part I), also showed that there were variations in the distribution patterns of papillae between children and adults. Most of the children (72.2%) had a cluster of fungiform papillae in the tip of the tongue and a decrease in density toward the posterior tongue. A similar distribution pattern was only observed with 36.7% of adults. Accordingly it is suggested that differences in the distribution patterns should be considered when comparisons of papillae densities are made between adults and children, especially when small areas on the tip of tongue are used to predict the total number of papillae on anterior tongue.

Practically, the variation in distribution patterns amongst individuals may explain classification of individuals based on their taste sensitivity to PROP. Currently, individuals are classified into 3 groups, namely, non-tasters, medium-tasters and super-tasters (Bartoshuk 1979; Bartoshuk et al., 1994). It has been reported that super-tasters of PROP have a higher papillae density on the anterior tongue than the non-tasters (Bartoshuk et al., 1994). Based on the distribution pattern, super-tasters may have a cluster of

156 Chapter 6 fungiform papillae on the anterior tongue whereas the papillae of medium- tasters and non-tasters may have a more uniform distribution. Further studies are required to investigate whether the distribution pattern affects the classification of individuals based on their taste sensitivity to PROP.

Interestingly, differences in distribution patterns could clarify other discrepancies in taste development studies. For example, it could explain why small areas of the anterior tongue of children produce higher local taste sensitivity (Stein et al., 1994). Stein studied the sensitivity of 12 small localised areas of the anterior tongue of adults and 8 year old males and reported areas closer to the tip of the tongue had higher sensitivity and more fungiform papillae than adults. However, according to a whole mouth study by (James et al., (1997) children (8-9 year old males) had a lower taste sensitivity than the adults when detecting sucrose, sodium chloride and citric acid. Overall the variation in distribution patterns of fungiform papillae may prove to be an important finding as regards taste perception differences during human development and between individuals.

An important finding in Chapter 3 (Part II) was that it demonstrated that it is possible to use a small area of the anterior tongue to predict fungiform papillae density of the whole anterior tongue (Correa et al., 2013). Previously a single small area was chosen without validation. The present study located an area that reliably predicted papillae density for 7-10 year olds (Area 1) and another for older children and adults (Area 5). Area 1 was in the tip of the anterior tongue, whereas, Area 5 was located in the mid-region of the anterior tongue. Together with development of the digital camera procedure the use of a single validated area provides a rapid and efficient procedure for counting fungiform papillae in humans (Correa et al., 2013).

In Chapter 3 (Part III), fungiform papillae counts were quantified within a 0.28 cm2 circle on the tip (Area X) and mid region (Area Y) of the tongue to investigate the validity of previous findings (Fast et al., 2002; Yackinous and Guinard 2002). The results showed that there was a significant relationship between the total number of fungiform papillae in the anterior two cm of the

157 Chapter 6 tongue and the 0.28 cm2 circle in both regions, when data for all participants were combined, validating the earlier findings. However, Areas X and Y produced less reliable densities than indicated for Areas 1 and 5 for particular age groups (Correa et al., 2013).

Following the above experiment further refinements of the digital camera procedure were investigated with the aim of using a more powerful high- resolution camera, which potentially could allow papillae counts without the use of any staining and drying methods. This was particularly relevant to measurements with patients with low tolerance to having ther tongue touched during papillae assessment.

Chapter 4 showed that by using a moderately high-resolution digital camera it is possible to measure papillae density without any disturbance to the surface of the tongue by dyes or drying the tongue (Correa et al., 2015) which have traditionally been used in such studies. A key outcome of this development was that it showed that the procedure could potentially be used with patients suffering from illnesses that cause ulceration or other lesions on the tongue and who would have found the staining and drying process painful and uncomfortable (e.g. xerostomia). The non-staining and non-drying procedures are also valuable methods for counting fungiform papillae in young children since it is simple and truly non-invasive. Additionally, the new no dye or drying method could be used with participants who are allergic to dyes commonly used in histology, hence increasing participant recruitment (Rosenhall 1982; Chadwick et al., 1990; Chung et al., 2001; Yusim et al., 2007; Aydogan et al., 2008).

The successful use of the high-resolution camera with the non-staining and non-drying procedures also made it possible to examine the vascular network of the fungiform papillae from the topographical images, which showed similar findings to those of Negoro et al., (2004), who used contact endoscopy to observe the vascular network of fungiform papillae.

158 Chapter 6

To date there have been limited anatomical studies of taste loss arising from drugs, other diseases or medical interventions. Clearly a limitation to anatomical studies has been the unsuitability of procedures, such as dissection of fungiform papillae from the surface of the tongue. The high- resolution camera, non-staining and non-drying procedures provide new methods for such studies.

Chapter 5 was a practical demonstration of the findings of Chapter 3 (Parts I to III). In Chapter 5 the high-resolution digital camera procedure which used the non-staining and non-drying method was utilized in a clinical situation, namely, to determine if the density of fungiform papillae in children with chronic kidney disease differed from that of clinical controls. The results showed that there was a significant decrease in the number of fungiform papillae in children with CRF in comparison to their clinical controls, and importantly there were also a significant correlation between the reduction of GFR and a decrease of fungiform papillae density (Correa et al., 2015).

The latter findings complement the report by Astback et al., (1999), who found that 78% of the fungiform papillae in CRF patients did not contain taste buds in comparison to only 12% of the papillae of controls. The present and past data, therefore, indicate that CRF patients have lower numbers of papillae and taste buds, which would lower the number of taste receptor cells within these entities and decrease their ability to perceive tastes. This is in agreement with a number of studies that have reported that adults suffering from renal failure have impaired taste acuity (Fernstrom et al., 1996; Middleton et al., 1999; Matson et al., 2003) and children with CRF have a poorer ability to perceive common tastes than their clinical controls (Armstrong et al., 2010). Overall, the latter procedure of measuring papillae density provides a simple way for following the loss and recovery of taste function in individual patients with CRF and other diseases that are known to affect taste function.

In conclusion, the digital camera-based method (Shahbake et al., 2005) can be used in a variety of studies for quantification of fungiform papillae on the

159 Chapter 6 anterior tongue. The study showed there were significant differences in the counts of fungiform papillae in 7-8 year olds and adults and that 11-12 year olds had similar counts to adults. In addition, there were differences in distributions of fungiform papillae between children and adults and there was a clear demonstration that it is possible to predict the relative level of fungiform papillae density of the whole anterior tongue by counting in a small region of the tongue in children and adults (Correa et al., 2013). On the other hand, non-staining and non-drying methods for fungiform papillae quantification were developed for clinical settings and it was utilised with patients with CRF to investigate the effect of renal disease on papillae density. Interestingly, there was a significant decrease of fungiform papillae counts in children with CRF in comparison with their clinical controls, suggesting taste loss could be the cause of malnutrition in some of the children (Correa et al., 2015). Accordingly, all the aims described in Chapter 1 were achieved.

160 References

References

Abdollahi, M. and Radfar, M. (2003). A review of drug-induced oral reactions. The Journal of Contemporary Dental Practice, 4(1): 10-31.

Agarwal, N., Phadke, K.D., Garg, I. and Alexander, P. (2003). Acute renal failure in children with idiopathic nephrotic syndrome. Pediatric Nephrology, 18(12): 1289-1292.

Akella, G.D., Henderson, S.A. and Drewnowski, A. (1997). Sensory acceptance of Japanese green tea and soy products is linked to genetic sensitivity to 6-n-propylthiouracil. Nutrition and Cancer, 29: 146-151.

Arey, L.B., Tremaine, M.J. and Mozingo, F.L. (1935). The numerical and topographical relations of taste buds to human circumvallate papillae throughout the life span. Anatomical Record, 64: 9-25.

Armstrong, J.E., Laing, D.G., Wilkes, F.J. and Kainer G. (2010). Smell and taste function in children with chronic kidney disease. Paediatric Nephrology 25:1497-504.

Arvidson, K. (1976). Scanning electron microscopy of fungiform papillae on the tongue of man and monkey. Acta Oto Laryngologica (stockh), 81: 496- 502.

Arvidson, K. (1979). Location and variation in number of taste buds in human fungiform papillae. Scandinavian Journal of Dental Research, 87: 435-442.

Arvidson, K. and Friberg, U. (1980). Human taste response and taste bud number in fungiform papillae. Science, 209: 807-808.

161 References

Astback, J., Fernstorm, A., Hylander, B., Arvidson, K. and Johansson, O. (1999). Taste buds and neuronal markers in patients with chronic renal failure. Peritoneal Dialysis International, 19: s315-s323.

Atkin-Thor, E., Goddard, B.W., O'Nion, J., Stephen, R.L. and Kolff, W.J. (1978). Hypogeusia and zinc depletion in chronic dialysis patients. The American Journal of Clinical Nutrition, 31: 1948-1951.

Aydogan, F., Celik, V., Uras, C., Salihoghu, Z. and Topuz, U. (2008). A comparison of the adverse reactions associated with isosulfan blue versus methylene blue dye in sentinel lymph node biopsy for breast cancer. American Journal of Surgery, 195(2): 277-278.

Ballard, C. (1997). How do we taste and smell? Hove, East Sussex: Wayland, 18-23.

Bartoshuk, L.M. (1979). Bitter taste of saccharin related to the genetic ability to taste the bitter substance 6-n-propylthiouracil. Science, 205(4409):934- 935.

Bartoshuk, L.M. (2000). Comparing sensory experiences across individuals: recent psychophysical advances illuminate genetic variation in taste perception. Chemical Senses, 25: 447-460.

Bartoshuk, L.M., Duffy, V.B. and Miller, I.J. (1994). PTC/PROP tasting: anatomy, psychophysics and sex effects. Physiology and Behavior, 56(6): 1165-1171.

Bartoshuk, L.M., Duffy, V.B., Reed, D. and Williams, A. (1996). Supertasting, earaches and head Injury: genetics and pathology alter our taste worlds. Neuroscience and Biobehavioral Reviews, 20: 79-87.

Bartoshuk, L.M., Gent, J., Catalanotto, F.A. and Goodspeed, R.B. (1983). Clinical evaluation of taste. American Journal of Otolaryngology, 4: 257-260.

162 References

Baumann, C.R., Ott, P.M. and Siegel, A.M. (2004). Hypogeusia as an adverse reaction of pheytoin. British Journal of Clinical Pharmacology, 58(6): 678-679.

Beauchamp, G. K., Bertino, M., Burke, D. and Engelman, K. (1990). Experimental sodium depletion and salt taste in normal human volunteers. The American Journal of Clinical Nutrition, 51: 881-889.

Beauchamp, G. K., Yamazaki, K. and Boyse, E. A. (1985). The chemosensory recognition of genetic individuality. Scientific American, 253: 86-92.

Beauchamp, G.K. and Moran, M. (1984). Acceptance of sweet and salty tastes in 2 year old children. Appetite, 5: 291-305.

Belecky, T.L. and Smith, D.V. (1990). Postnatal development of palatal and laryngeal taste buds in the hamster. Journal of Comparative Neurology, 293(4): 646-654.

Bell, G., W., Large, D., M. and Barclay, S., C. (2000). Oral health care in diabetes mellitus. Journal of the South African Dental Association, 55:158-65.

Bhatia, S. and Sharma, K.N. (1991). Taste impairment for glucose in diabetic PTC tasters and non-tasters. Diabetes Research and Clinical Practice, 12: 193-200.

Blakeslee, A.F. (1932). Genetics of sensory thresholds: taste for phenylthiocarbamide. Proceedings of the National Academy of Sciences of USA, 18: 120-130.

Blanchart, A., De Carlos J.A. and López-Mascaraque, L. (2006). Time frame of mitral cell development in the mice olfactory bulb. The Journal of Comparative Neurology, 496(4):529-543.

163 References

Blendis, L.M., Ampil, M., Wilson, D.R., Kiwan, J., Labranche, J., Johnson, M. and Williams, C. (1981). The importance of dietary protein in the zinc deficiency of uremia. The American Journal of Clinical Nutrition, 34: 2658- 2661.

Bohatyrewicz, M., Borowiecka, E. and Ciechanowski, K. (2005). Taste disorders present in patients with chronic renal failure do not have an effect on overhydration and excessive weight gain between dialysis. Nephrology (Carlton), 10(3): 274-275.

Bolze, M.S., Fosmire, G.J., Stryker, J.A., Chung, C.K. and Flipse, B.G. (1982). Taste acuity, plasma zinc levels and weight loss during radiotherapy: a study of relationships. Radiology, 144: 163-169.

Bonanni, G. and Perazzi, F. (1965). Behavior of taste sensitivity in patients subjected to high energy radiologic treatment for tumors of the oral cavity. Nuntius Radiologicus, 31(4): 383-397.

Bradley, R.M. (1971). Tongue topography. In: Beidler, L.M., (Ed.) Handbook of Sensory Physiology, Vol.4 (Part 2, Chapter 1), Springer-Verlag, N.Y.

Bradley, R.M. (1972). Development of the taste bud and gustatory papillae in human foetuses. In: Bosma, J.F. and Charles, C. (Eds.). The Mouth of the Infant; Third Symposium on Oral Sensation and Perception, Springfield IL, 137-161.

Bradley, R.M. and Stern, I.B. (1967). The development of the human taste bud during the foetal period. Journal of Anatomy, 101(4): 743-752.

Bradley, R.M., Stedman, H.M. and Mistretta, C.M. (1985). Age does not affect numbers of taste buds and papillae in adults Rhesus monkey. The Anatomical Record, 212: 246-249.

Brand, J.G. (1989). Receptor events and transduction in taste and olfaction, N.Y., M. Dekker Publication, 31-40.

164 References

Breen, F.M., Plomin, R. and Wardle, J. (2006). Heritability of food preferences in young children. Physiology and Behavior, 88(4-5): 443-447.

British Renal Foundation (BRF) (2006). Chronic kidney disease in adults; UK guidelines for identification, management and referral. Royal College of Physicians of London. 3-11.

Burge, J.C., Park, H.S., Whitlock, C.P. and Schemmel, R.A. (1979). Taste acuity in patients undergoing long term hemodialysis. Kidney International, 15: 49-53.

Burkl, W. (1954). Ueber das vorkommen von geschmacksknospen im mittleren drittel des oesophagus. Anatomische Anzeiger, 100: 320-321.

Cabral, P.C., Diniz-Ada, S. and de Arruda, I.K. (2005). Vitamin A and zinc status in patients on maintenance haemodialysis. Nephrology (Carlton), 10(5): 459-463.

Caring for Australians with renal impairment: The CARI guidelines (2006). www.kidney.org.au/cari/guidelines.php. Accessed 14 August 2007.

Carl, W. (1988). Managing the oral manifestations of cancer therapy, Part 1: Head and neck radiation therapy. Compendium, 9(4): 306-314.

Cassileth, B.R., Lusk, E.J., Strouse, T.B., Miller, D.S., Brown, L.L., Cross, P.A. and Tenaglia, A.N. (1984). Psychosocial status in chronic illness. A comparative analysis of six diagnostic groups. The New England Journal of Medicine, 311(8): 506-511.

Chadwick, B.L., Hunter, M.L., Evans, M.T. and Hunter, B. (1990). Allergic reaction to food dye patent blue. British Dental Journal, 168(10): 386-387.

Cheng, K.K.F. (2007). Oral mucositis, dysfunction and distress in patients undergoing cancer therapy. Journal of Clinical Nursing, 16(11): 2114-2121.

165 References

Cheng, L.H. and Robinson, P.P. (1991). The distribution of fungiform papillae and taste buds on the human tongue. Archives of Oral Biology, 36: 583-589.

Chung, K., Baker, J.R., Baldwin, J.L. and Chou, A. (2001). Identification of carmine allergens among three carmine allergy patients. Allergy, 56(1): 73- 77.

Ciechanover, M., Peresecenschi, G., Aviram, A. and Steiner, J.E. (1980). Malrecognition of taste in uremia. Nephron, 26: 20-22.

Ciges, M. and Morales, J. (1980). Experimental ageusia. Acta Otolaryngol, 89: 240-248.

Coe, K. and Lail, C. (2007). Peritoneal dialysis in the neonatal intensive care unit. Management of acute renal failure after a severe subgaleal haemorrhage. Advances in Neonatal Care, 7(4): 179-186.

Cohen, A.M. and Vig, P.S. (1976). A serial growth study of the tongue and intermaxillary space. The Angle Orthodontist, 46(4): 332-337.

Coleman, J.E., Watson, A.R., Chowdhury, S., Thurlby, D. and Wardell, J. (2002). Comparison of two micronutrient supplements in children with chronic renal failure. Journal of Renal Nutrition, 12(4): 244-247.

Collin, H., Niskanen, N., Uusitupa, M., Töyry, J., Collin, P., Koivisto, A., Viinamäki, H. and Meurman, J.H. (2000). Oral symptoms and signs in elderly patients with type 2 diabetes mellitus; A focus on diabetic neuropathy. Oral Surgery, Oral Medicine, Oral Pathology, 90: 299-305.

Combarros, O., Miró, J. and Berciano, J. (1994). Ageusia associated with thalamic plaque in multiple sclerosis. European Neurology, 34(6):344-346.

Conger, A.D. (1973). Loss and recovery of taste acuity in patients irradiated to the oral cavity. Radiation Research, 53: 338-347.

166 References

Correa, M., Hutchinson, I., Laing, D.G. and Jinks, A.L. (2013). Changes in fungiform papillae density during development in humans. Chemical Senses, 38: 519-27.

Correa, M., Laing, D.G., Hutchinson, I., Jinks, A.L., Armstrong, J.E. and Kainer, G. (2015). Reduced taste function and taste papillae density in children with chronic kidney disease. Pediatric Nephrology, 30(11):2003- 2010, DOI: 10.1007/s00467-015-3131-5 (Epub 2015 June 5).

Cowart, B.J. (1981). Development of taste perception in humans: sensitivity and preferences throughout the life span. Psychological Bulletin, 90: 43-73.

Crook, C.K. and Lipsitt, L.P. (1976). Neonatal nutritive sucking: effects of taste stimulation upon sucking rhythm and heart rate. Child Development, 47: 518-522.

Crook, C.K., (1978). Taste perception in the newborn Infant. Behaviour and Development, 1: 52-69.

Deems, R.O. and Friedman, M.I. (1988). Sodium chloride preference is altered in a rat model of liver disease. Physiology and Behavior, 43(4): 521- 525.

Delay, R.J., Kinnamon, J.C. and Roper, S.D. (1986). Ultrastructure of mouse vallate taste buds. II. Cell types and cell lineage. Journal of Comparative Neurology, 253: 242-252.

Desor, J.A. and Beauchamp, G.K. (1987). Longitudinal changes in sweet preferences in humans. Physiology and Behavior, 39(5):639-641.

Desor, J.A. and Maller, O. (1975). Taste correlates of disease states: Cystic fibrosis. The Journal of Pediatrics, 87: 93-96.

167 References

Desor, J.A., Maller, O. and Andrews, K. (1975). Ingestive responses of human newborns to salty, sour, and bitter stimuli. Journal of Comparative and Physiological Psychology, 89(8): 966-970.

DeWys, W.D. (1974). Abnormalities of taste as a remote effect of a neoplasm (part VIII: a spectrum of organ systems that respond to the presence of cancer). Annals New York Academy of Sciences, 230: 427-434.

Dobell, E., Chan, M., Williams, P. and Allman, M. (1993). Food preferences and food habits of patients with chronic renal failure undergoing dialysis. Journal of the American Dietetic Association, 93(10): 1129-1135.

Douma, C.E. and Smit, W. (2006). When to start dialysis? Nephrology, Dialysis, Transplantation, 21(supplement 2): ii20-24.

Duffy, V.B., Peterson, J.N., Dinehart, M.E. and Bartoshuk, L.M. (2003). Genetic and environmental variation in taste: association with sweet intensity, preference and intake. Topics in Clinical Nutrition, 18: 209-220.

Dvornik, S., Cuk, M., Racki, S. and Zaputovic, L. (2006). Serum zinc concentrations in the maintenance of hemodialysis patients. Collegium Antropologicum, 30(1): 125-129.

Emura, S., Hayakawa, D., Chen, H. and Shoumura, S. (2001). Morphology of the dorsal lingual papillae in the newborn panther and Asian black bear. Okajimas Folia Anatomica Japonica, 78: 173-178.

Esses, B.A., Jafek, B.W., Hommel, D.J. and Eller, P.M. (1988). Histological and ultrastructural changes of the murine taste bud following ionizing irradiation. Ear, Nose and Throat Journal, 67(7): 478-493.

Essick, G.K., Chopra, A., Guest, S. and McGlone, F. (2003). Lingual tactile acuity, taste perception and the density and diameter of fungiform papillae in female subjects. Physiology and Behavior, 80(2-3): 289-302.

168 References

Farbman, A.I. (1980). Renewal of taste bud cells in rat circumvallate papillae. Cell and Tissue Kinetics, 13: 349-357.

Farbman, A.I. (1990). Olfactory neurogenesis: genetic or environmental controls? Trends in Neurosciences, 32: 489-502.

Fast, K., Duffy, V.B. and Bartoshuk, L.M. (2002). New psychophysical insight in evaluating genetic variation in taste In: Rouby, C., Schaal, B., Dubois, D., Gervais, R. and Holey, A. (eds.), Olfaction, Taste and Cognition, Cambridge University Press, N.Y., 391-407.

Fernstorm, A., Hylander, B. and Rossner, S. (1996). Taste acuity in patients with chronic renal failure. Clinical Nephrology, 45(3): 169-174.

Finger, T.E. and Silver W.L. (1991). Neurobiology of taste and smell. Malabar, Fla: Krieger Publication.

Fowler, E.B., Breault, L.G. and Cuenin, M.F. (2001). Periodontal disease and its association with systemic disease. Military Medicine, 166(1):85-89.

Fox, N.A., (1985). Sweet/sour interest/disgust The role of approach- withdrawal in the development of emotions. In: Fox (Ed.), Social perception in infants, Norwood, NJ Ablex, pp.53-71.

Gadoth, N., Margalith, D., Schlien, N., Svetliza, E. and Litwin, A. (1982). Presence of fungiform papillae classic dysautonomia. The Johns Hopkins Medical Journal, 151: 298-301.

Ganchrow, D, Ganchrow, J.R., Verdin-Alcazar, M. and Whitehead, M.C. (2003a). Brain derived neurotrophic factor, neurotrophin-3 and Tyrosine Kinase receptor-like immunoreactivity in lingual taste bud fields of mature hamster. The Journal of Comparative Neurology, 455: 11-24.

Ganchrow, D., Ganchrow, J.R., Verdin-Alcazar, M. and Whitehead, M.C. (2003a). Brain derived neurotrophic factor, neurotrophin-3 and Tyrosine

169 References

Kinase receptor-like immunoreactivity in lingual taste bud fields of mature hamster. The Journal of Comparative Neurology, 455: 11-24.

Ganchrow, D., Ganchrow, J.R., Verdin-Alcazar, M. and Whitehead, M.C. (2003b). Brain derived neurotrophic factor, neurotrophin-3 and Tyrosine Kinase receptor-like immunoreactivity in lingual taste bud fields of mature hamster after sensory denervation. The Journal of Comparative Neurology, 455: 25-39.

Ganchrow, J.R. Steiner, J.E. and Mumif, D. (1983). Neonatal facial expressions in response to different qualities and intensities of gustatory stimuli. Infant Behavior and Development, 6:473-484.

Gardiner, J., Barton, D., Vanslambrouck, J.M., Braet, F., Hall, D., Marc, J. and Overall, R. (2008). Defects in tongue papillae and taste sensation indicate a problem with neurotrophic support in various neurological diseases. The Neuroscientist, 14: 240-250.

Gartner, L.P. and Hiatt, J.L. (2000). Color atlas of histology. Philadelphia, London : Lippincott Williams & Wilkins.

Gedikli, O., Doğru, H., Aydin, G., Tüz, M., Uygur, K. and Sari, A. (2001). Histopathological changes of chorda tympani in chronic otitis media. The Laryngoscope, 111:724-727.

Ghazali, S. and Barratt, T.M. (1974). Urinary excretion of calcium and magnesium in children. Archives of Disease in Childhood, 49: 97-101.

Goldstein, J. (1989). Sensation and perception. Wadsworth Publishing Company, Belmont, California, 513-523.

Guthrie, K.M., Wilson, D.A. and Leon, M. (1990). Early unilateral deprivation modifies olfactory bulb function. The Journal of Neuroscience, 10(10):3402- 3412.

170 References

Guyton, A.C. (1987). Basic Neuroscience. In: Anatomy and Physiology, WB Saunders Co., USA, 321-327.

Hambidge, K.M., Hambidge, C., Jacobs, M. and Baum, J.D. (1972). Low levels of zinc in hair, anorexia, poor growth, and hypogeusia in children. Pediatric Research, 6(12):868-874.

Heiderich, F. (1906). Die Zahl und die dimension der geschmacksknospen der papilla vallata des menschen in den verschiedenen lebensaltern. Nachr. Koen. Gesell. Wissensch. Goettingenetic Math Physik, K.L., 54-64.

Hendricks, J., Calasanti, T.M. and Turner, H.B. (1988). Food ways of the elderly. American Behavioral Scientist, 32(1): 61-68.

Henkin, R.I. and Powell, G.F. (1962). Increased sensitivity of taste and smell in cystic fibrosis. Science, 138:1107-1108.

Henkin, R.I. and Shallenberger, R.S. (1970). Aglycogeusia: the inability to recognize sweetness and its possible molecular basis. Nature, 227: 965-966.

Henkin, R.I., Schechter, P.J., Hoye, R. and Mattern, C.F.T. (1971). Idiopathic hypogeusia with dysgeusia, hyposmia and dysosmia; a new syndrome. Journal of American Medical Record Association, 217(4): 434-440.

Henkin, R.I., Talal, N., Larson, A.L. and Mattern, C.F.T. (1972). Abnormalities of taste and smell in Sjogren's syndrome. Annals of Internal Medicine, 76(3): 375-383.

Hermal, J., Schonwetter, S.V. and Samueloff, S. (1970). Taste sensation and age in man. Journal of Oral Medicine, 25: 39-42.

Hersch, M. and Ganchrow, D. (1980). Scanning electronmicroscopy of developing papillae on the tongue of human embryos and fetuses. Chemical Senses, 5: 331-341.

171 References

Hertz, J., Cain, W.S., Bartoshuk, L.M. and Dolan, T.F. (1975). Olfactory and taste sensitivity in children with cystic fibrosis. Physiology and Behavior, 14(1): 89-94.

Hill, D.L. (2001). Taste development. In, Handbook of Behavioural Neurobiology 13, Developmental Psychobiology, Blass, E. (Ed.). Kluwer Academic/ Plenum Publications, N.Y., 517-549.

Hinds, J.W. and Hinds, P.L. (1976). Synapse formation in the mouse olfactory bulb. I. Quantitative studies. The Journal of Comparative Neurology, 169(1):15-40.

Hofer, H.O. (1977). Taste areas of the plica sublingualis of Alouatta and Aotus (Primates, Platyrrhini). Cell and Tissue Research, 177: 415-429.

Hou-Jensen, H. (1933). Die papillae foliata des menschen. Zeitschrift fur Anatomica Entwicklungs Geschichte, 102: 348-388.

Huang, M.C., Chen, M.E., Hung, H.C., Chen, H.C., Chang, W.T., Lee, C.H., Wu, Y.Y., Chiang, H.C. and Hwang, S.J. (2008). Inadequate energy and excess protein intakes may be associated with worsening renal function in chronic kidney disease. Journal of Renal Nutrition, 18(2): 187-194.

Hurley, R.S., Herbert, L.A. and Rypien, A.B. (1987). A comparison of taste acuity for salt in renal patients vs. normal subjects. Journal of the American Dietetic Association, 87(11): 1531-1534.

Iida, M., Yoshioka, I. and Muto, H. (1983). Taste bud papillae on the retromolar mucosa of the rat, mouse and golden hamster. Acta Anatomica, 117: 374-381.

James, C.E., Laing, D.G. and Oram, N. (1997). A comparison of the ability of 8-9 year old children and adults to detect taste stimuli. Physiology and Behavior, 62(1): 193-197.

172 References

Jung, H.S., Akita, K. and Kim, J.Y. (2004). Spacing patterns on tongue surface-gustatory papilla. International Journal of Developmental Biology, 48 (2-3): 157-161.

Just, T., Pau, H.W., Bombor, I., Guthoff, R.F., Fietkau, R. and Hummel, T. (2005). Confocal microscopy of the peripheral gustatory system: Comparison between healthy subjects and patients suffering from taste disorders during radiochemotherapy. The Laryngoscope, 115(12): 2178-2182.

Just, T., Pau, H.W., Witt, M. and Hummel, T. (2006). Contact endoscopic comparison of morphology of human fungiform papillae of healthy subjects and patients with transected chorda tympani nerve. The Laryngoscope, 116: 1216-1222.

Kalmus, H. (1976). PTC testing of infants. Annals of Human Genetics, 40: 139-140.

Kho, H.S., Lee, S.W., Chung, S.C. and Kim, Y.K. (1999). Oral manifestations and salivary flow rate, pH and buffer capacity in patients with end-stage renal disease undergoing hemodialysis. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontics, 88(3): 316-319.

Kiewe, P., Jovanovic, S., Thiel, E. and Korfel, A. (2004). Reversible ageusia after chemotherapy with pegylated liposomal doxorubicin. The Annals of Pharmacotherapy, 38: 1212-1214.

Ko, K.W. (1992). Haemorrhagic fever with renal syndrome: clinical aspects. Pediatric Nephrology, 6(2): 197-200.

Koehler, J. and Leonhaeuser, I.U. (2008). Changes in food preferences during aging. Annals of Nutrition and Metabolism, 52(1): 12-14.

Koivisto, U.K. and Sjoden, P.O. (1996). Reasons for rejection of food items in Swedish families with children aged 2-17. Appetite, 26: 89-103.

173 References

Krimm, R.F., Miller, K.K., Kitzman, P.H., Davis, B.M. and Albers, K.M. (2001). Epithelial over-expression of BDNF or NT4 disrupts targeting of taste neurons that innervate the anterior tongue. Developmental Biology, 232: 508- 521.

Kullaa-Mikkonen, A. and Sorvari, T.E. (1985). A scanning electron microscopic study of the dorsal surface of the human tongue. Acta Anatomica (Basel), 123: 114-120.

Kullaa-Mikkonen, A., Koponen, A. and Seilonen, A. (1987). Quantitative study of human fungiform papillae and taste buds: variation with aging in different morphological forms of the tongue. Gerodontics, 3: 131-135.

Kutuzov, H. and Sicher, H. (1951). The filiform and the conical papillae of the tongue in the white rat. The Anatomical Record, 110: 275-288.

Laing, D.G., Armstrong, J.E., Aitken, M., Carroll, A., Wilkes, F.J., Jinks, A.L. and Jaffé, A. (2010). Chemosensory function and food preferences of children with cystic fibrosis. Pediatric Pulmonology, 45(8):807-815.

Laing, D.G., Segovia, C., Fark, T., Laing, O.N., Jinks, A.L., Nikolaus, J. and Hummel, T. (2008). Tests for screening olfactory and gustatory function in school-age children. Otolaryngology and Head and Neck Surgery, 139(1): 74-82.

Lalonde, E.R. and Eglitis, J.A. (1961). Number and distribution of taste buds on epiglottis, pharynx, larynx, soft palate and uvula in human newborn. The Anatomical Record, 140: 91-93.

Landis, B.N., Beutner, D., Frasnelli, J., Huttenbrink, H., and Hummel, T. (2005). Gustatory function in chronic inflammatory middle ear diseases. The Laryngoscope, 115:1124–1127.

Laville, M. and Fouque, D. (2000). Nutritional aspects in hemodialysis. Kidney International Supplement, 76: S133-139.

174 References

Lawson, W.B., Zeidler, A. and Ruberstein, A.H. (1979). Taste detection and preference in diabetics and their relatives. Psychosomatic Medicine, 41: 219- 220.

Lemon, C.H. and Katz, D.B. (2007). The neural processing of taste. BMC Neuroscience, 8: S3-5.

Leshem, M., Rudoy, J. and Schulkin, J. (2003). Calcium taste preference and sensitivity in humans. II. Hemodialysis patients. Physiology and Behavior, 78(3): 409-414.

Lindemann, B. (1996). Taste reception. Physiological Reviews, 76(3): 719- 766.

Liu, S., Manson, J. E., Lee, I. M., Cole, S. R., Hennekens, C. H., Willett, W. C. and Buring, J. E. (2000). Fruit and vegetable intake and risk of cardiovascular disease: the Women's Health Study. The American Journal of Clinical Nutrition, 72: 922-928.

Lobstein, T., Baur, L. and Uauy, R. (2004). Obesity in children and young people: A crisis in public health. Obesity Reviews, 5: 4-84.

Lozada-Nur, F., Cram, D. and Gorsky, M. (1992). Clinical response to levamisole in thirty-nine patients with erythema multiforme. Oral Surgery, Oral Medicine and Oral Pathology, 74(3): 294-298.

Mafra, D., Cuppari, L. and Cozzolino, S.M. (2002). Iron and zinc status of patients with chronic renal failure who are not on dialysis. Journal of Renal Nutrition, 12(1): 38-41.

Magarey, A., Daniels, L. A. and Smith, A. (2001). Fruit and vegetable intakes of Australians aged 2-18 years: an evaluation of the 1995 National Nutrition Survey data. Australian and New Zealand Journal of Public Health, 25: 155- 161.

175 References

Malauiya, G.N. and Ramu, G. (1981). Loss of taste and somatic sensations over the tongue in leprous facial palsy – a case report. Leprosy in India, 53(4): 656-659.

Manlapas, F.C., Stein, A.A., Pagliara, A.S., Apicelli, A.A., Porter, B.A., and Patterson, P.R. (1965). Phenylthiocarbamide taste sensitivity in cystic fibrosis, The Journal of Pediatrics, 66: 8-11.

Martini, F.H. (2006). The Oral Cavity, In: Fundamentals of Anatomy & Physiology 7th edition. Prentice Hall, New Jersey, 853-860.

Matson, A., Wright, M., Oliver, A., Woodrow, G., King, N., Dye, L., Blundell, J., Brownjohn, A. and Turney, J. (2003). Zinc supplementation at conventional doses does not improve the disturbance of taste perception in hemodialysis patients. Journal of Renal Nutrition, 13(3): 224-228.

Matsukawa, N. and Okada, S. (1994). Microvasculature of the lingual papillae in primates and insectivores fungiform, vallate and foliate papillae. Okajimas Folia Anatomica Japonica, 71(4): 259-277.

Mbiene, J.P., Maccallum, D.K. and Mistretta, C.M. (1997). Organ cultures of embryonic rat tongue support tongue and gustatory papilla morphogenesis in vitro without intact sensory ganglia. The Journal of Comparative Neurology, 377: 324-340.

Mennella, J. A., Pepino, M. Y. and Beauchamp, G. K. (2003). Modification of bitter taste in children. Developmental Psychobiology, 43:120-127.

Mennella, J.A., Kennedy, J.M. and Beauchamp, G.K. (2006). Vegetable acceptance by infants: effects of formula flavours. Early Human Development, 82(7): 463-468.

Middleton, R.A. and Allman-Farinelli, M.A. (1999). Taste sensitivity is altered in patients with chronic renal failure receiving continuous ambulatory peritoneal dialysis. The Journal of Nutrition, 129: 122-125.

176 References

Miller, I.J. (1986). Variation in human fungiform taste bud densities among regions and subjects. Anatomical Record, 216: 474-482.

Miller, I.J. (1987). Human fungiform taste bud density and distribution. In: Roper, S.D. and Atema, J. Annals New York Academy of Science, Vol.510: Olfaction and Taste IX, The New York Academy of Science, N.Y., 501-503.

Miller, I.J. (1988a). Human taste bud density across adult age groups. Journal of Gerontology, 43(1): B26-30.

Miller, I.J. (1988b). Spatial features of taste receptor populations. In: Miller, I.J. (Ed.). The Beidler Symposium on Taste and Smell, Book Series Association, Florida, 45-55.

Miller, I.J. (1995). Anatomy of the peripheral taste system. In: Doty, R.L. (Ed.). Handbook of Olfaction and Gustation, Marcel Dekker Inc., N.Y., 521- 547.

Miller, I.J. and Bartoshuk, L.M. (1991). Taste perception, taste bud distribution and spatial relationship. In: Getchell, T.V., Doty, R.L., Bartoshuk, L.M. and Snow, J.B. (Eds.). Smell and Taste; Health and Disease, Raven Press, N.Y., 205-233.

Miller, I.J. and Preslar, A.J. (1974). Spatial distribution of rat fungiform papillae. Anatomical Record, 181: 679-684.

Miller, I.J. and Reedy, F.E. (1988). Human taste pore quantification with video-microscopy. Chemical Senses, 13: 719-725.

Miller, I.J. and Reedy, F.E. (1990a). Quantification of fungiform papillae and taste pores in living human subjects. Chemical Senses, 15: 281-294.

Miller, I.J. and Reedy, F.E. (1990b). Variation in human taste bud density and taste sensitivity. Physiology and Behavior, 47(6): 1213-1219.

177 References

Miller, I.J. and Smith, D.V. (1984). Quantitative taste bud distribution in hamster. Physiology and Behavior, 32(2): 275-285.

Mistretta, C.M. (1972). Topographical and histological study of the developing rat tongue, papillae and taste buds. In: Bosma, J.F. and Charles, C. (Ed.). The Mouth of the Infant, Third Symposium on Oral Sensation and Perception, Springfield IL., 163-187.

Mistretta, C.M. (1984). Aging effects on anatomy and neurophysiology of taste and smell. Gerodontology, 3: 131-136.

Mistretta, C.M. (1991). Developmental neurobiology of taste system. In: Getchell, T.V., Doty, R.L., Bartoshuk, L.M. and Snow, J.B. (Ed.). Smell and Taste; Health and Disease, Raven Press, N.Y., 35-47.

Mistretta, C.M. and Baum, B.J. (1984). Quantitative study of taste buds in fungiform and circumvallate papillae of young and aged rats. Journal of Anatomy, 138: 323-332.

Mistretta, C.M. and Oakley, L.A. (1986). Quantitative anatomical study of taste buds in fungiform papillae of young and old Fischer rats. Journal of Gerontology, 41(3): 315-318.

Mistretta, C.M., and Hill, D.L. (2003). Development of the taste system: basic neurobiology. In Doty, R.L. (Ed.), Handbook of Olfaction and Gustation, 2nd edition. Marcel Dekker, N.Y., 759–782.

Mistretta, C.M., and Liu, H. (2006). Development of fungiform papillae: Patterned lingual gustatory organs. Archives of Histology and Cytology, 69(4): 199-208.

Mistretta, C.M., Gurkan, S. and Bradley, R.M. (1988). Morphology of chorda tympani fiber receptive fields and proposed neural rearrangements during development. The Journal of Neuroscience, 8(1):73-78.

178 References

Miyata, T., Wada, Y., Cai, Z., Iida, Y., Horie, K., Yasuda, Y., Maeda, K. and Kurokawa, K. (1997). Implication of an increased oxidative stress in the formation of advanced glycation end products in patients with end-stage renal failure. Kidney International, 51(4): 1170-1181.

Mochizuki, Y. (1937). An observation on the numerical and topographical relations of taste buds to circumvallate papillae of Japanese. Okajimas Folia Anatomica Japonica, 15: 595-608.

Mochizuki, Y. (1939). Studies on the papilla foliata of Japanese (I) The number of papillae foliatae (II) The number of taste buds. Okajimas Folia Anatomica Japonica, 18: 337-369.

Moore, L.M., Nielsen, C.R. and Mistretta, C.M. (1982). Sucrose taste thresholds: age-related differences. Journal of Gerontology, 37(1): 64-69.

Moreau, R. and Lebrec, D. (2007). Diagnosis and treatment of acute renal failure in patients with cirrhosis. Best Practice and Research. Clinical Gastroenterology, 21(1): 111-123.

Moses, S.W., Rotem, Y., Jagoda, N., Talmor, N., Eichhorn, F. and Levin, S. (1967). Clinical genetic and biochemical study of familial dysautonomia in Israel. Israel Journal of Medical Science, 3: 358-371.

Mossman, K.L. (1982). Quantitative radiation dose-response relationships for normal tissues in man (I) Gustatory tissue response during photon and neutron radiotherapy. Radiation Research, 91: 265-274.

Mossman, K.L. (1983). Quantitative radiation dose-response relationships for normal tissue in man. II. Response of the salivary glands during radiotherapy. Radiation Research, 95: 392-398.

Mossman, K.L. and Henkin, R.I. (1978). Radiation induced changes in taste acuity in cancer patients. International Journal of Radiation, Oncology, Biology and Physics, 4: 663-670.

179 References

Mossman, K.L., Chencharick, J.D., Scheer, A.C., Walker, W.P., Ornitz, R.D., Rogers, C.C. and Henkin, R.I. (1979). Radiation-induced changes in gustatory function comparison of effects of neutron and photon irradiation. International Journal of Radiation, Oncology, Biology and Physics, 5: 521- 528.

Murphy, C., Schubert, C. R., Cruickshanks, K. J., Klein, B. E., Klein, R. D. and Nondahl, M. (2002). Prevalence of olfactory impairment in older adults. The Journal of the American Medical Association, 288: 2307-2312.

Murray, R.G. and Murray, A. (1971). Relations and possible significance of taste bud cells. Contributions to Sensory Physiology, 5: 47-95.

Must, A. and Strauss, R., S. (1999) Risks and consequences of childhood and adolescent obesity. International Journal of Obesity, 23: S2–S11.

Nagy, J.I., Goedert, M., Hunt, S.P. and Bond, A. (1982). The nature of the substance P-containing nerve fibres in taste papillae of the rat tongue. Neuroscience, 7(12):3137-51.

Nakashima, T., Kimmelman, C. P. and Snow, J. B. (1984). Structure of human fetal and adult olfactory neuroepithelium. Archives of Otolaryngology, 110: 641-646.

Negoro, A., Umemoto, M., Fukazawa, K., Terada, T. and Sakagami, M. (2004). Observation of tongue papillae by video-microscopy & contact endoscopy to investigate their correlation with taste function. Auris Nasus Larynx, 31(3): 255-259.

Nelson, G.M. (1998). Biology of taste buds and the clinical problem of taste loss. The Anatomical Record, 253(3): 70-78.

Nicklaus, S., Boggio, V. and Issanchou, S. (2005). Gustatory perceptions in children. Archives de Pediatrie, 12(5): 579-584.

180 References

Nilsson, B. (1979). The occurrence of taste buds in the palate of human adults as evidenced by light microscopy. Acta Odontological Scandinavica, 37: 253-258.

Nin, T., Sakagami, M., Sone-Okunaka, M., Muto, T., Mishiro, Y. and Fukazawa, K. (2006). Taste function after section of chorda tympani nerve in middle ear surgery. Auris Nasus Larynx, 33: 13-17.

North America Kidney Disease Outcome Quality Intuitive: National Kidney Foundation. (2002). K/DOQI Clinical Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification and Stratification. American Journal of Kidney Disease, 39: S1-S266.

Oakley, B. (1988). Taste bud development in rat vallate and foliate papillae. In Hink, P., Soukup, T., Vejsada, R. and Zelena, J. (Eds.). Mechanoreceptor, Plenum, N.Y., 17-22.

Oakley, B. and Witt, M. (2004). Building sensory receptors on the tongue. Journal of Neurocytology, 33: 631-646.

Oakley, B., LaBelle, D.E., Riley, R.A., Wilson, K. and Wu, L.H. (1991). The rate and locus of development of rat vallate taste buds. Developmental Brain Research, 58: 215-221.

Ochsenbein-Kolble, N., Mering, R.V., Zimmermann, R. and Hummel, T. (2005). Changes in gustatory function during the course of pregnancy and postpartum. An International Journal of Obstetrics and Gynaecology, 112: 1636-1640.

Oh, K.J., Lee, H.H., Lee, J.S., Chung, W., Lee, J.H., Kim, S.H. ands Lee, J.S. (2006). Reversible renal vasoconstriction in a patient with acute renal failure after exercise. Clinical Nephrology, 66(4): 297-301.

181 References

Oram, N., Laing, D.G., Freeman, M.H. and Hutchinson, I. (2001). Analysis of taste mixtures by adults and children. Developmental Psychobiology, 38(1):67-77.

Osaki, T., Ohshima, M., Tomita, Y. Matsugi, N. and Nomura, Y. (1996). Clinical and physiological investigations in patients with taste abnormality. Journal of Oral Pathology and Medicine, 25(1): 38-43.

Paran, N., Mattern, C.F.T. and Henkin, R.I. (1975). Ultrastructure of the taste bud of the human fungiform papilla. Cell Tissue Research, 161: 1-10.

Peck, C.C. and Hannam, A.G. (2007). Human jaw and muscle modelling. Archives of Oral Biology, 52(4): 300-304.

Petruzzi, M., Bendittis, M.D., Grassi, R., Cassano, N., Vena, G. and Serpico, R. (2002). Oral lichen planus: a preliminary clinical study on treatment with tazarotene. Oral Diseases, 8: 291-295.

Plattig, K.H. (1984). The sense of taste. In: Piggott, J.R., Sensory Analysis of Foods, 2nd edition, Elsevier Applied Sciences, London, N.Y., 1-23.

Pljesa, S., Golubovic, G., Tomasevic, R., Markovic, R. and Perunicic, G. (2005). “Watermelon stomach” in patients on chronic hemodialysis. Renal Failure, 27(5): 643-646.

Prutkin, J., Fisher, E.M., Etter, L., Fast, K., Gardner, E., Lucchina, L.A., Snyder, D.J., Tie, K., Weiffenbach, J. and Bartoshuk, L.M. (2000). Genetic variation and inferences about perceived taste intensity in mice and men. Physiology and Behavior, 69(1-2): 161-173.

Rakic, P. and Riley, K.P. (1983). Overproduction and elimination of retinal axons in the fetal rhesus monkey. Science, 219(4591):1441-1444.

Rao, G.N. and Sisodia, P. (1970). Diabetes and PTC tasting ability. Journal of Associate Physiology (India), 18: 57.

182 References

Reiter, E.R., DiNarda, L.J. and Costanzo, R.M. (2006). Toxic effects on gustatory function. Advances in Oto-Rhino-Laryngology, 63: 265-277.

Richter, C.P. and Campbell, H.K. (1940). Sucrose taste thresholds of rats and humans. American Journal of Physiology, 128: 291-297.

Robinson, P.P. and Winkles, P.A. (1990). Quantitative study of fungiform papillae and taste buds on the cat's tongue. The Anatomical Record, 225: 108-111.

Rosenhall, L. (1982). Evaluation of intolerance to analgesics, preservatives and food colorants with challenge tests. European Journal of Respiratory Diseases, 63(5): 410-419.

Rosenstein, D. and Oster, H. (1988). Differential facial responses to four basic tastes in newborns. Child Development, 59: 1555-1568.

Rupesh, S. and Nayak, U.A. (2007). Genetic sensitivity to the bitter taste of 6-n propylthiouracil: a new risk determinant for dental caries in children. Journal of the Indian Society of Pedodontics and Preventive Dentistry, 24(2):63-68.

Sandow, P.L., Hejrat-Yazdi, M. and Heft, M.W. (2006). Taste loss and recovery following radiation therapy. Journal of Dental Research, 85(7): 608- 611.

Sarafidis, P.A., Li, S., Chen, S.C., Collins, A.J., Brown, W.W., Klag, M.J. and Bakris, G.L. (2008). Hypertension awareness, treatment and control in chronic kidney disease. The American Journal of Medicine, 121(4): 332-340.

Schieber, F. (1992). Aging and the senses. In Birren, J.E., Sloane, R.B. and Cohen, G.D., (Eds.), Handbook of mental health and aging, New York: Academic Press, 251-306.

183 References

Schiffman, S.S. (1993). Perception of taste and smell in elderly persons. Critical Reviews in Food Science and Nutrition, 33(1): 17-26.

Schmidt, F. and Thews, G. (1980). Human physiology, Springer-Verlag Berlin Heidelberg, N.Y., 300-306.

Schwartz, G.J., and Furth, S.L. (2007). Glomerular filtration rate measurement and estimation in chronic kidney disease. Pediatric Nephrology, 22(11): 1839-1848.

Segovia, C., Hutchinson, I., Laing, D.G. and Jinks, A.L. (2002). A quantitative study of fungiform papillae and taste pore density in adults and children. Developmental Brain Research, 138: 135-146.

Shahbake, M. (2002). Taste bud density in children with different taste sensitivities. Honours thesis, University of Western Sydney.

Shahbake, M., Hutchinson, I., Laing, D.G. and Jinks, A.L. (2005). Rapid quantitative assessment of fungiform papillae density in the human tongue. Brain Research, 1052: 196-201.

Sharaby, A.E., Ueda, K., Honma, S. and Wakisaka, S. (2006). Initial innervation of the palatal gustatory epithelium in the rat as revealed by growth-associated protein-43 (GAP-43) immunohistochemistry. Archives of Histology and Cytology, 69(4): 257-272.

Shatzman, A.L. and Henkin, R.I. (1981). Gustin concentration changes relative to salivary zinc and taste in humans. Proceedings of the National Academy of Sciences of the United States of America, 78(6): 3867-3871.

Sicher, H. (1966). Obran’s Oral Histology and Embryology. 6th edition. St. Louise, The C.V. Mosby Company. 256-263.

Smith, A., Farbman, A. and Dancis, J. (1965). Absence of taste bud papillae in Familial dysautonomia. Science, 47(3661): 1040-1041.

184 References

Sone, M., Hayashi, H., Yamamoto, H., Tominaga, M. and Nakashima, T. (2003). A comparative study of intratympanic steroid and no synthase inhibitor for treatment of cochlear lateral wall damage due to acute otitis media. European Journal of Pharmacology, 15: 313-318.

Sone, M., Sakagami, M., Tsuji, K. and Mishiro, Y. (2001). Younger patients have a higher rate of recovery of taste function after middle ear surgery. Archives of Otolaryngology, Head and Neck Surgery, 127:967-969.

Stein, N., Laing, D.G. and Hutchinson, I. (1994). Topographical differences in sweetness sensitivity in the peripheral gustatory system of adults and children. Developmental Brain Research, 82: 286-292.

Sveja, J. and Janota, M. (1972). Papillae foliatae der menschlichen zunge. Deutsche zahnärztliche Zeitschrift, 27: 685-692.

Svejda, J. and Janota, M. (1974). Scanning electron microscopy of the papillae foliatae of the human tongue. Oral Surgery, Oral Medicine and Oral Pathology, 37: 208-216.

Teixeira-Nunes, F., de Campos, G., Xavier de Paula, S.M., Merhi, V.A., Portero-McLellan, K.C., da Motta, D.G. and de Oliveira, M.R. (2008). Dialysis adequacy and nutritional status of hemodialysis patients. Hemodialysis International, 12(1): 45-51.

Temple, E.C. Hutchinson, I., Laing, D.G. and Jinks, A.L. (2002). Taste development: differential growth rates of tongue regions in humans. Developmental Brain Research, 135: 65-70.

Tepper, B.J. (1998). 6-n-Propylthiouracil: a genetic marker for taste, with implications for food preference and dietary habits. American Journal of Human Genetics, 63(5):1271-1276.

Tepper, B.J. and Nurse, R.J. (1997). Fat perception is related to PROP taster status. Physiology and Behavior, 61(6): 949-954.

185 References

Tepper, B.J., Banni, S., Melis, M, Crnjar, R. and Barbarossa, I.T. (2014). Genetic sensitivity to the bitter taste of 6-n-Propylthiouracil (PROP) and its association with physiological mechanisms controlling Body Mass Index (BMI). Nutrient, 6: 3363-3381.

Terry, M.C. and Segall, G. (1947). The Association of diabetes and taste blindness. The Journal of Heredity, 38: 136-138.

Trof, R.J., Di Maggio, F., Leemreis, J. and Groeneveld, A.B. (2006). Biomarkers of acute renal injury and renal failure. Shock, 26(3): 245-253.

Umemoto, M., Sone, M., Fujii, M. and Sakagami, M. (1999) The change of fungiform papillae of tongue after chorda tympani damage: long term observation. Japanese Journal of Taste and Smell Research, 6: 525–526.

Vissink, A., Burlage, F.R., Spijkervet, F.K.L., Jansma, J. and Coppes, R.P. (2003a). Prevention and treatment of the consequences of head and neck radiotherapy. Critical Reviews in Oral Biology and Medicine, 14(3): 213-225.

Vissink, A., Jansma, J., Spijkervet, F.K.L., Burlage, F.R. and Coppes, R.P. (2003b). Oral sequelae of head and neck radiotherapy. Critical Reviews in Oral Biology and Medicine, 14(3): 199-212.

Wardle, L., Cooke, L.L., Gibson, E.L., Sapochnik, M., Sheiham, A. and Lawson, M. (2003). Increasing children’s acceptance of vegetables; a randomized trial of parent-led exposure. Appetite, 40: 155-162.

Weiffenbach, J.M., Baum, B.J. and Burghauser, R. (1982). Taste thresholds: quality specific variation with human aging. Journal of Gerontology, 37(3): 372-377.

Weiler, E. and Farbman, A. I. (1997). Proliferation in the rat olfactory epithelium: Age-dependent changes. Journal of Neuroscience, 17: 3610- 3622.

186 References

Whissell-Buechy, D. (1990). Effects of age and sex on taste sensitivity to PTC in Berkeley Guidance sample. Chemical Senses, 15: 39-57.

Whitehead, M.C., Beeman, C.S. and Kinsella, B.A. (1985). Distribution of taste and general sensory nerve endings in fungiform papillae of the hamster. The American Journal of Anatomy, 173(3):185-201.

Wilson, J.G. (1905). The structure and function of taste buds of larynx. Brain, 28: 339-351.

Witt, M. and Kasper, M. (1999). Distribution of cytokeratin filaments and vimentin in developing human taste buds. Anatomy and Embryology, 199: 291-299.

Witt, M. and Reutter, K. (1997). Scanning electron microscopical studies of developing gustatory papillae in humans. Chemical Senses, 22: 601-612.

Wöckel, L., Jacob, A., Holtmann, M. and Poustk, F. (2008). Reduced number of taste papillae in patients with eating disorders. Journal of Neural Transmission, 115: 537-544.

Wotman, S., Mandel, I.D., Khotim, S. (1964). Salt taste thresholds and cystic fibrosis, American Journal of Diseases of Children, 108:372-374.

Yackinous, C. A. and Guinard, J.X. (2002). Relation between PROP (6-n- propylthiouracil) taster status, taste anatomy and dietary intake measures for young men and women. Appetite, 38:201-209.

Yamashita, H., Nakagawa, K., Nakamura, N., Abe, K., Asakage, T., Ohmoto, M., Okada, S., Matsumoto, I., Hosoi, Y., Sasano, N., Yamakawa, S. and Ohtomo, K. (2006a). Relation between acute and late irradiation impairment of four basic tastes and irradiated tongue volume in patients with head-and- neck cancer. International Journal of Radiation Oncology, Biology, Physics, 66(5): 1422-1429.

187 References

Yamashita, H., Nakagawa, K., Tago, M., Nakamura, N., Shiraishi, K., Eda, M., Nakata, H., Nagamatsu, N., Yokoyama, R., Onimura, M. and Ohtomo, K. (2006b). Taste dysfunction in patients receiving radiotherapy. Head and Neck, 28: 508-516.

Yusim, Y., Livingstone, D. and Sidi, A. (2007). Blue dyes, blue people: the systemic effects of blue dyes when administered via different routes. Journal of Clinical Anesthesia, 19(4): 315-321.

Zhang, G., Zhang, H., Wang, X., Zhan, Y., Deng, S. and Qin, Y. (2009). The relationship between fungiform papillae density and detection threshold for sucrose in young males. Chemical Senses, 34: 93-99.

Zheng, W.K., Inokuchi, A., Yamamoto, T. and Komiyama, S. (2002). Taste dysfunction in irradiated patients with head and neck cancer. Fukuoka Igaku Zasshi, 93(4): 64-76.

Zuniga, J., Davis, S., Englehatd, R., Miller I.J., Schiffman, S. and Phillips, C. (1993). Taste performance on the anterior human tongue varies with fungiform taste bud density. Chemical Senses, 18: 449-460.

188 Appendix A

Appendix A

Table A1: Descriptive data of children and adult groups Subject number Male/Female Age Fungiform papillae Count / 0.28 cm2 Digital camera Videomicroscope 1 M 8 16 14 2 M 9 13 12 3 M 9 12 10 4 F 8 17 11 5 F 8 14 15 6 F 9 18 18 7 F 9 10 13 8 F 9 16 12 9 F 9 19 19 11A M 25 6 6 12 F 28 14 9 13 F 33 11 14 14 F 38 7 8 15 F 32 13 8 16 F 31 10 9 17 F 30 11 14 Total Mean - 12.94 12.00 S. E. - 0.93 0.91 Mean (Children) 8.67 15.00 13.78 S. E. 0.17 0.99 1.02 Mean (Adult) 31.00 10.29 9.71 S. E. 1.54 1.10 1.17

Table A2: A 2X2 ANOVA between factors digital camera and video- microscopy method and within factors adult versus children (Method X Age)

Type III Sum of Mean Source Squares df Square F Sig. Method X Age 159.556 3 53.185 5.853 0.003 Method 6.334 1 6.334 0.697 0.411 Age 151.691 1 151.691 16.695 0.000 Error 254.413 28 9.086

189 Appendix A

Table A3: A Single factor ANOVA for participants with large differences in fungiform papillae counts between digital camera and video-microscopy method (i.e. participant 4, 8, 12 and 15).

Type III Sum of Mean Source Squares df Square F Sig. Between 50 1 50.00 15 0.008 methods Within Methods 20 6 3.33 Total 70 7

190 Appendix B

Appendix B

Table B1: Example of the worksheet used for recording the fungiform papillae number in each computerized grid Date:______Guide lines measurements

Subject#:______Picture#:______Scale=___ Grid No. of grids (stained area photographed)= ______=_____Cm 2cm=____Grid

Grids

Cm Total 1 2

25mm2 Overall

Note:______

191 Appendix B

Table B2: Descriptive data of 7-8 year olds Subject # F/ M Fungiform papillae density Type Fungiform papillae A/B density (0.28 cm2 circle) Total First cm Second Area X Area Y cm 1 F 167 112 55 A 44 19 2 M 180 116 64 A 47 23 3 F 129 112 17 A 42 24 4 F 177 129 48 A 45 15 5 M 116 85 31 B 40 17 6 M 178 126 52 B 37 19 7 F 171 125 46 A 46 13 8 M 141 86 55 B 50 13 9 F 133 84 49 A 37 17 10 F 174 95 79 B 34 22 11 M 140 77 63 A 48 15 12 F 148 96 52 A 32 19 13 F 191 127 64 A 45 16 14 F 191 119 72 B 33 19 15 F 142 102 40 A 29 16 16 M 175 115 60 B 32 16 17 M 195 105 90 A 45 24 18 F 169 116 53 A 36 14 19 F 178 116 62 A 39 17 20 F 162 113 49 A 49 21 21 F 156 123 33 A 37 15 22 F 160 110 50 A 32 12 23 F 190 112 78 A 44 22 24 F 158 121 37 A 45 28 25 M 170 117 53 B 41 22 26 M 147 107 40 B 29 20 27 M 168 129 39 A 27 13 28 M 147 102 45 A 40 20 29 M 152 111 41 A 35 15 Mean (total) 162.24 109.93 52.31 39.31 18.14 S.E. 3.74 2.63 2.89 1.22 0.74 Mean (female) 164.47 112.47 52.00 39.35 18.18 S.E. 4.59 2.97 3.88 1.47 1.02 Mean (male) 159.08 106.33 52.75 39.25 18.08 S.E. 6.41 4.73 4.52 2.17 1.10 Minimum 116 77 17 27 12 Maximum 195 129 90 50 28

192 Appendix B

Table B3: Descriptive data of 9-10 year olds Subject # F/ M Fungiform papillae density Type Fungiform papillae A/B density (0.28 cm2 circle) Total First cm Second Area X Area Y cm 1 F 161 89 72 A 45 24 2 M 149 85 64 B 53 25 3 F 123 89 34 A 25 18 4 M 115 66 49 B 26 16 5 F 152 95 57 A 49 17 6 M 140 81 59 A 44 18 7 M 148 107 41 A 31 18 8 M 75 52 23 B 39 16 9 M 159 95 64 A 52 22 10 F 135 75 60 A 44 18 11 M 148 98 50 A 34 20 12 M 159 93 66 B 35 24 13 M 165 108 57 A 36 16 14 M 129 79 50 B 51 19 15 M 126 83 43 B 31 11 16 F 180 136 44 A 39 13 17 M 131 63 68 B 42 19 18 M 134 77 57 B 40 14 19 M 167 128 39 A 35 19 20 M 157 119 38 A 38 12 21 F 186 125 61 A 42 17 22 F 158 105 53 A 33 14 23 F 163 110 53 A 34 16 Mean (Total) 146.09 93.87 52.22 39.04 17.65 S.E. 4.69 3.09 2.43 1.62 0.77 Mean (female) 157.38 103.00 54.25 40.00 17.73 S.E. 8.33 7.15 4.05 1.86 1.07 Mean (male) 140.07 88.93 51.2 37.25 17.50 S.E. 5.21 5.42 3.28 3.19 1.04 Minimum 75 52 23 25 11 Maximum 186 136 72 53 25

193 Appendix B

Table B4: Descriptive data of 11-12 year olds Subject # F/ M Fungiform papillae density Type Fungiform papillae A/B density (0.28 cm2 circle) Total First cm Second Area X Area Y cm 1 M 107 72 35 A 33 11 2 M 93 66 27 B 36 8 3 M 145 89 56 A 31 17 4 F 178 139 39 A 59 14 5 M 138 79 59 A 34 16 6 M 188 107 81 A 52 26 7 M 190 112 78 A 47 25 8 M 152 105 47 A 41 18 9 F 112 65 47 A 34 16 10 F 111 82 29 B 35 16 11 M 165 113 52 A 40 7 12 F 85 51 34 B 24 9 13 M 146 100 46 A 45 16 14 M 141 86 55 A 35 17 15 F 146 83 63 A 39 17 16 F 156 84 72 A 42 19 17 F 217 165 52 A 53 15 18 F 158 93 65 A 43 16 19 M 156 95 61 A 45 15 20 M 172 103 69 A 37 12 21 F 164 81 83 B 39 14 22 F 137 74 63 A 27 15 23 M 129 54 75 A 29 13 24 M 160 111 49 A 22 14 25 F 203 128 75 B 49 26 26 F 182 99 83 B 52 19 27 M 158 117 41 B 30 21 28 M 131 82 49 B 32 18 29 F 111 70 41 A 35 17 30 F 150 87 63 B 34 18 31 F 151 68 83 A 33 20 32 M 174 113 61 A 47 29 33 F 170 109 61 A 30 21 Mean (Total) 150.79 93.39 57.39 38.30 16.82 S.E. 5.27 4.24 2.81 1.54 0.86 Mean (female) 151.94 92.38 59.56 39.25 17.00 S.E. 8.81 7.46 4.41 2.46 0.93 Mean (male) 149.71 94.35 55.35 37.41 16.65 S.E. 6.27 4.54 3.58 1.92 1.46 Minimum 85 51 27 22 7 Maximum 217 165 83 59 29

194 Appendix B

Table B5: Descriptive data Adults Subject # F/ M Fungiform papillae density Type Fungiform papillae A/B density (0.28 cm2 circle) Total First cm Second Area X Area Y cm 1 F 107 102 5 A 33 11 2 F 161 83 78 A 46 13 3 F 180 114 66 A 31 17 4 F 111 77 34 B 36 14 5 F 183 118 65 A 34 16 6 M 173 87 86 B 57 26 7 F 163 114 49 A 36 25 8 F 157 108 49 A 39 18 9 F 156 108 48 A 34 16 10 F 101 64 37 B 35 16 11 F 156 128 28 B 32 17 12 F 222 160 62 B 54 14 13 F 124 75 49 B 36 16 14 F 93 56 37 A 35 17 15 F 106 70 36 B 39 17 16 F 147 102 45 B 42 19 17 M 114 70 44 B 43 15 18 F 129 87 42 B 33 16 19 F 124 76 53 B 35 15 20 F 133 93 34 B 37 12 21 F 141 93 48 A 39 14 22 M 133 88 83 B 47 15 23 F 192 109 53 B 39 23 24 M 124 75 49 B 37 14 25 F 113 63 50 B 38 22 26 M 147 103 44 B 38 19 27 M 136 81 55 B 30 14 28 M 119 83 36 B 32 18 29 M 148 88 60 A 35 17 30 M 113 71 42 A 34 18 Mean (Total) 140.20 91.30 48.90 37.87 16.80 S.E. 5.50 4.07 3.01 1.14 0.64 Mean (female) 141.95 95.24 47.57 38.19 16.76 S.E. 7.11 5.48 3.74 1.47 0.84 Mean (male) 134.11 82.89 51.22 37.11 16.89 S.E. 8.25 3.43 4.95 1.77 0.89 Minimum 93 56 5 30 11 Maximum 222 160 86 57 26

195 Appendix B

Table B6: A 2X4 ANOVA main effect (groups; 7-8 year olds to adult) and between factor (sex) and total fungiform papillae density in the first 2 cm

Sum of Source Squares df Mean Square F Sig. Group X Sex 2077.510 7 692.503 0.974 0.408 Sex 23.817 1 23.817 0.033 0.855 Group 6907.545 3 2302.515 3.238 0.025 Error 76095.822 107 711.176 Total 2672373.000 115

Table B7: Post-hoc Tukey HSD paired comparison analysis between all age groups

Source Mean Difference Standard Error P

7-8 yr X 9-10 yr 16.15 7.45 0.138 7-8 yr X 11-12 yr 11.45 6.79 0.335 7-8 yr X adults 22.04 6.95 0.010 9-10 yr X 11-12 yr 4.70 7.24 0.916 9-10 yr X adults 5.89 7.39 0.856 11-12 yr X adults 10.59 6.73 0.398

Table B8: A 2X4 ANOVA main effect (groups; 7-8 year olds to adult) and between factor (sex) and fungiform papillae density in the first cm interval

Sum of Source Squares df Mean Square F Sig. Group X Sex 8234.305 7 1176.329 2.993 0.007 Group 6226.726 3 2075.575 5.282 0.002 Sex 8.914 1 8.194 0.021 0.885 Error 42049.225 107 392.983 Total 1134842.000 115

196 Appendix B

Table B9: Post-hoc Tukey HSD paired comparison analysis between all age groups in the first cm interval

Source Mean Difference Standard Error P

7-8 yr X 9-10 yr 16.062 5.535 0.023 7-8 yr X 11-12 yr 16.537 5.046 0.008 7-8 yr X adults 18.631 5.162 0.003 9-10 yr X 11-12 yr 0.476 5.385 0.979 9-10 yr X adults 2.570 5.494 0.966 11-12 yr X adults 2.094 5.001 0.975

Table B10: A 2X4 ANOVA main effect (groups; 7-8 year olds to adult) and between factor (sex) and fungiform papillae density in the second cm interval

Sum of Source Squares df Mean Square F Sig. Group X Sex 1356.586 7 193.789 0.804 0.586 Group 931.120 3 310.373 1.288 0.282 Sex 59.952 1 59.952 0.249 0.619 Error 25791.188 107 241.039 Total 348489.000 115

197 Appendix B

Table B11: Fungiform papillae density of the 8 small areas in the first 2 cm of the tongue in 7-8 year olds

Total papillae Fungiform papilla count in small area Subject # density 1 2 3 4 5 6 7 8 1 167 42 30 16 24 9 22 3 21 2 180 50 20 20 26 20 12 16 16 3 129 54 26 15 17 6 3 5 3 4 177 58 24 22 25 16 12 10 10 5 116 44 17 14 10 14 8 6 3 6 178 63 24 23 16 17 14 12 9 7 171 57 16 20 32 19 10 7 10 8 141 47 13 16 10 13 14 14 14 9 133 43 12 18 11 17 12 8 12 10 174 46 15 20 14 19 20 21 19 11 140 27 13 20 17 13 17 12 21 12 148 35 16 27 18 12 17 10 13 13 191 50 22 27 28 24 15 12 13 14 191 51 21 29 18 19 16 20 17 15 142 52 18 15 17 12 14 7 7 16 175 51 17 23 24 17 10 18 15 17 195 37 27 23 18 21 19 26 24 18 169 50 21 24 21 14 21 9 9 19 178 62 20 25 9 12 29 17 4 20 162 43 21 29 20 12 10 13 14 21 156 52 24 28 19 11 12 4 6 22 160 47 18 29 16 17 11 13 9 23 190 40 26 24 22 22 22 23 11 24 158 53 23 26 19 12 10 7 8 25 170 48 21 29 19 19 10 10 14 26 147 41 19 19 28 11 12 8 9 27 168 47 33 24 25 12 7 7 13 28 147 45 21 17 19 17 6 5 17 29 152 42 22 27 20 14 8 7 12 Mean 162.24 47.48 20.69 22.38 19.38 15.21 13.55 11.38 12.17 Error 3.74 1.45 0.92 0.89 1.05 0.78 1.05 1.10 0.99 Minimum 116 27 12 14 9 6 3 3 3 Maximum 195 63 33 29 32 24 29 26 24

198 Appendix B

Table B12: Fungiform papillae density of the 8 small areas in the first 2 cm of the tongue in 9-10 year olds Fungiform papilla count in small area Subject # Total papillae density 1 2 3 4 5 6 7 8 1 161 44 13 20 12 21 17 15 19 2 149 39 10 20 16 21 18 16 9 3 123 31 12 24 22 11 8 11 4 4 115 22 10 18 16 13 18 9 9 5 152 43 19 18 15 14 14 15 14 6 140 28 21 8 24 11 15 9 24 7 148 33 23 22 29 16 6 3 16 8 75 20 7 11 14 9 7 4 3 9 159 31 16 26 22 19 19 17 9 10 135 36 13 15 11 13 26 4 17 11 148 39 22 20 17 15 7 5 23 12 159 30 21 24 18 20 11 20 15 13 165 47 27 16 18 21 12 13 11 14 129 39 9 15 16 8 21 4 17 15 126 29 16 14 24 9 18 4 12 16 180 47 29 31 29 11 14 11 8 17 131 25 11 13 14 11 13 21 23 18 134 39 10 13 15 12 16 12 17 19 167 43 25 31 29 5 10 7 17 20 157 44 22 28 25 8 11 8 11 21 186 47 25 26 27 16 14 16 15 22 158 36 20 25 24 15 17 14 7 23 163 46 20 26 18 20 17 12 4 Mean 146.09 36.43 17.43 20.17 19.78 13.87 14.30 10.87 13.22 Error 4.97 1.72 1.34 1.34 1.18 0.99 1.03 1.13 1.27 Minimum 75 20 7 8 11 5 6 3 3 Maximum 186 47 29 31 29 21 26 21 24

199 Appendix B

Table B13: Fungiform papillae density of the 8 small areas in the first 2 cm of the tongue in 11-12 year olds Fungiform papilla count in small area Subject # Total papillae density 1 2 3 4 5 6 7 8 1 107 29 10 15 18 11 6 8 10 2 93 34 8 6 18 4 10 2 11 3 145 32 17 16 24 13 18 14 11 4 178 47 29 34 29 15 9 7 8 5 138 32 10 13 24 14 12 18 15 6 188 42 24 23 18 21 19 27 14 7 190 47 22 19 24 23 21 23 11 8 152 42 25 20 18 15 10 9 13 9 112 29 9 11 16 23 9 7 8 10 111 29 18 17 18 12 10 4 3 11 165 42 27 16 28 16 17 12 7 12 85 15 10 10 16 7 11 6 10 13 146 40 17 18 25 14 15 10 7 14 141 33 16 17 20 20 16 11 8 15 146 42 10 15 16 17 19 12 15 16 156 34 13 21 16 18 20 19 15 17 217 58 37 39 31 15 10 12 15 18 158 43 22 15 13 16 21 15 13 19 156 43 19 14 19 24 12 14 11 20 172 42 24 20 17 24 22 11 12 21 164 37 15 16 13 23 22 27 11 22 137 39 13 12 10 12 22 19 10 23 129 19 11 15 9 21 19 20 15 24 160 41 27 23 20 20 13 9 7 25 203 47 31 23 27 24 17 19 15 26 182 37 23 20 19 25 20 22 16 27 158 35 27 30 25 14 12 9 6 28 131 31 14 20 17 12 19 9 9 29 111 29 15 17 9 15 7 7 12 30 150 37 17 14 19 20 12 17 14 31 151 25 13 15 15 24 19 22 18 32 174 44 29 20 20 17 13 14 17 33 170 41 26 23 19 22 17 21 1 Mean 150.79 36.88 19.03 18.39 19.09 17.30 15.12 13.82 11.15 Error 5.27 1.50 1.31 1.14 0.95 0.92 0.84 1.14 0.69 Minimum 85 15 8 6 9 4 6 2 1 Maximum 217 58 37 39 31 25 22 27 18

200 Appendix B

Table B14: Fungiform papillae density of the 8 small areas in the first 2 cm of the tongue in adults

Total papillae Fungiform papilla count in small area Subject # density 1 2 3 4 5 6 7 8 1 107 42 29 16 15 2 1 1 1 2 161 28 22 10 23 21 33 16 8 3 180 44 25 17 28 27 24 8 7 4 111 11 13 20 33 17 13 2 2 5 183 41 27 20 30 24 17 13 11 6 173 38 23 13 13 27 35 12 12 7 163 35 39 26 14 11 16 10 12 8 157 51 17 19 21 13 16 7 13 9 156 41 28 19 20 14 15 7 12 10 101 22 15 11 16 9 9 11 8 11 156 59 33 20 16 7 6 7 8 12 222 94 23 20 23 16 20 16 10 13 124 33 15 15 12 16 12 4 17 14 93 15 4 14 23 11 17 7 2 15 106 27 5 11 27 7 17 7 5 16 147 39 17 19 27 15 17 10 3 17 114 32 15 10 13 15 12 8 9 18 129 42 12 15 18 12 12 8 10 19 124 32 21 13 10 14 15 8 11 20 133 35 18 17 23 16 16 4 4 21 141 42 16 19 16 15 11 8 14 22 133 27 16 28 17 14 12 12 7 23 192 45 16 25 23 20 30 16 17 24 124 37 11 19 8 14 17 10 8 25 113 22 10 13 18 14 9 15 12 26 147 44 19 22 18 15 20 5 4 27 136 42 11 13 15 19 14 14 8 28 119 32 14 15 22 10 15 5 6 29 148 28 17 19 24 13 17 14 16 30 113 20 13 17 21 12 14 11 5 Mean 140.20 36.67 18.13 17.17 19.57 14.67 16.07 9.20 8.73 Error 5.50 2.75 1.42 0.83 1.10 1.00 1.30 0.76 0.81 Minimum 93 11 4 10 8 2 1 1 1 Maximum 222 94 39 28 33 27 35 16 17

201 Appendix B

Table B15: Pearson Correlation analysis of total fungiform papillae density and fungiform papillae density in Areas X and Y of all age groups

Total FP vs. Area X Total FP vs. Area Y Pearson Correlation 0.491 0.325 Sig. (2-tailed) 0.000 0.000 N 115 115

Table B16: A 2X4 ANOVA main effect (groups; 7-8 year olds to adult) and between factor (sex) and fungiform papillae density of Area X

Sum of Source Squares df Mean Square F Sig. Group X Sex 67.608 7 22.536 0.398 0.755 Group 42.689 3 14.230 0,251 0.860 Sex 44.372 1 44.372 0.783 0.378 Error 6062.789 107 56.662 Total 177480.000 115

Table B17: A 2X4 ANOVA main effect (groups; 7-8 year olds to adult) and between factor (sex) and fungiform papillae density of Area Y

Sum of Source df Mean Square F Sig. Squares Group X Sex 46.892 7 6.699 0.077 0.972 Group 31.293 3 10.431 0.594 0.620 Sex 3.797 1 3.797 0.216 0.643 Error 1877.839 107 17.550 Total 36395.000 115

202 Appendix C

Appendix C

Table C1: Fungiform papillae density of 0.28 cm2 circle in 3 stages

Fungiform Papillae density/ 0.28 cm2 Subject # F/M Stage 1 Stage 2 Stage 3 1 M 19 17 20 2 M 13 13 15 3 F 13 17 21 4 F 12 12 15 5 F 6 6 8 6 M 9 8 11 7 F 14 14 14 8 F 11 16 16 9 F 13 14 13 10 M 11 10 14 11 M 11 9 13 12 F 14 14 12 13 F 12 10 11 14 F 7 7 7 15 F121212 16 M 11 11 11 17 F 10 10 10 18 F 13 12 13 19 F 9 9 10 20 F 14 14 16 21 F5 6 8 22 F 13 13 13 23 M 8 8 10 24 F 3 3 3 25 F 8 9 9 26 M 10 10 10 27 F7 7 7 28 F 17 16 20 29 F 17 17 17 30 F 11 11 12 31 F 10 10 11 32 F 13 12 14 33 F 9 12 12 34 M 8 9 9 35 F 10 7 8 Mean - 10.94 11 12.14 S.E. - 0.57 0.59 0.66 Minimum - 333 Maximum - 19 17 21

203 Appendix C

Table C2: A Bland–Altman (B-A) difference plot analysis showing 95 % Limits of Agreement (LoA) between the papillae densities of stage 1 and 2.

95% Confidence Interval of the Parameter Estimate S.E. Difference Mean difference -0.0571 -0.5837 to 0.4694 0.26

95% Lower LoA -3.06 -3.97 to -2.15 0.45

95% Upper LoA 2.94 2.04 to 3.86 0.45

Table C3: A Bland–Altman (B-A) difference plot analysis showing 95 % Limits of Agreement (LoA) between the papillae densities of stage 1 and 3.

95% Confidence Interval of the Parameter Estimate S.E. Difference Mean difference -1.20 -1.85 to 0.55 0.32

95% Lower LoA -4.94 -6.07 to -3.81 0.56

95% Upper LoA 2.54 1.41 to 3.67 0.56

Table C4: A Bland–Altman (B-A) difference plot analysis showing 95 % Limits of Agreement (LoA) between the papillae densities of stage 2 and 3.

95% Confidence Interval of the Parameter Estimate S.E. Difference Mean difference -1.14 -1.67 to -0.61 0.25

95% Lower LoA -4.16 -5.07 to -3.24 0.45

95% Upper LoA 1.87 0.96 to 2.78 0.45

Table C5: Linear regression analysis difference of fungiform papillae counts between stage 1 vs 2, stage 1 vs 3 and stage 2 vs 3.

Stage R R2 F t p Stage 1 vs. 0.052 0.003 0.90 -0.299 0.766 2 Stage 1 vs 3 0.277 0.077 2.74 -1.654 0.108 Stage 2 vs 3 0.287 0.083 2.97 -1.724 0.094

204 Appendix D

Appendix D

Table D1: Fungiform papillae count and eGFR score of each patient and their matched controls

Control Patient Subject # eGFR eGFR FP/ 0.28 cm2 FP/ cm2 FP/ 0.28 cm2 FP/ cm2 mL/min/1.73m2 mL/min/1.73m2 1 12 42.45 161 4 14.15 56.7 2 15 53.06 141 9 31.84 58 3 2 7.07 109 6 21.22 20 4 5 17.69 94 6 21.22 34 5 10 35.37 127 4 14.15 32 6 11 38.91 104 7 24.76 25 7 14 49.52 97 5 17.69 47 8 8 28.30 101 5 17.69 11.6 9 7 24.76 112 5 17.69 15 10 8 28.30 174 5 17.69 9.6 11 7 24.76 125 6 21.22 14 12 12 42.45 113 7 24.76 23 Mean 9.25 32.72 121.50 5.75 20.34 28.83 S.E. 1.095 3.87 7.36 0.41 1.45 4.92 Medium 9 31.84 112.5 5.5 19.46 24 Minimum 2 7.07 94 4 14.15 9.6 Maximum 15 53.06 174 9 31.84 58

205 Appendix D

Table D2: Compare group Kruskal- Wallis’ test between Fungiform papillae density of patient group and their matched controls

Kruskal-Wallis’ Group df n p Mean rank Statistic (X2) 7.169 Control 1 12 0.0074 16.33

- Patient - 12 - 8.67

Table D3: Compare group Kruskal- Wallis’ test between eGFR of patient group and their matched controls

Kruskal-Wallis’ Group df n p Mean rank Statistic (X2)

Control 1 12 0.000 18.50 17.280

Patient - 12 - 6.50 -

Table D4: Compare group post hoc Mann-Whitney test between Fungiform papillae density of patient group and their matched controls

Mann-Whitney Group n p (2-tailed) Mean rank Sum of rank Wilcoxon W Z statistic Statistic (U)

Control 12 0.0068 16.33 196.00 26.0 104.00 -2.68

Patient 12 - 8.67 104.00 - - -

Table D5: Compare group post hoc Mann-Whitney test between eGFR of patient group and their matched controls

Mann-Whitney Group n p (2-tailed) Mean rank Sum of rank Wilcoxon W Z statistic Statistic (U)

Control 12 0.000 18.50 222.00 0.00 78.00 -4.16

Patient 12 - 6.50 78.00 - - -

Table D6: Pearson correlation analysis between fungiform papillae count and eGFR score of the patient and their matched controls

Total FP vs. eGFR Pearson Correlation 0.562 p (2-tailed) 0.004 N 24

206 Paper 1: Shahbake et al., 2005 Appendix E

Brain Research 1052 (2005) 196 – 201 www.elsevier.com/locate/brainres Research Report Rapid quantitative assessment of fungiform papillae density in the human tongue

Maryam Shahbakea, Ian Hutchinsona, David G. Lainga,b,*, Anthony L. Jinksa

aChildren’s Food Research and Education Unit, Centre for Advanced Food Research, College of Science, Technology and Environment, University of Western Sydney, Locked Bag 1797, Penrith South, NSW 1797, Australia bSchool of Women and Children’s Health, Faculty of Medicine, University of New South Wales, Level 3, Sydney Children’s Hospital, High Street, Randwick, NSW 2031, Australia

Accepted 10 June 2005 Available online 26 July 2005

Abstract

Fungiform papillae density, which can be used in a variety of circumstances as an indicator of taste function [L.M. Bartoshuk, V.B. Duffy, I.J. Miller, PTC/PROP tasting: anatomy, psychophysics and sex effects, Physiol. Behav. 56 (1994) 1165–117; I.J. Miller, F.E. Reedy, Variation in human taste bud density and taste intensity perception, Physiol. Behav. 47 (1990) 1213–1219; J.R. Zuniga, N. Chen, C.L. Phillips, Chemosensory and somatosensory regeneration after lingual nerve repair in humans, J. Oral Maxillofac. Surg. 55 (1997) 2–13], was measured on the dorsal surface of the anterior tongue of living humans using a digital camera and a videomicroscope. Both procedures provided similar results, with the camera providing a more rapid, portable and flexible imaging procedure. Subsequently, the camera was successfully used to identify small regions of the anterior tongue which provide reliable measures of fungiform papillae density that correlate highly with the total number of fungiform papillae on the anterior tongue. D 2005 Elsevier B.V. All rights reserved.

Theme: Development and regeneration Topic: Sensory systems

Keywords: Fungiform papillae; Papillae density; Humans; Digital camera; Videomicroscopy

1. Introduction scopy [7–9,11]. However, although the videomicroscope is an excellent tool for this purpose, its use is limited to the High numbers of fungiform papillae on the anterior research laboratory. Currently, there is a need for a more dorsal surface of the tongue are commonly found in people portable system that allows filming of fungiform papillae of who are classified as supertasters of the bitter substance subjects at their bedside and outpatient clinics in hospitals to PROP compared to moderate and non-tasters [1,8], while gain an insight to taste function and of substantial numbers degeneration or loss of taste following medications or neural of children in schools in studies of the development of taste damage is accompanied by a decrease in the number of [10]. Another disadvantage of the videomicroscope is that it papillae found in individuals [11,15]. Measurement of requires 30–60 min to obtain high quality images from an papillae number or density, therefore, can provide informa- individual to allow counting of papillae. This time period is tion about taste function. Counts of these papillae in living unacceptable to patients in pain or uncomfortable from tissue have been achieved using non-invasive videomicro- clinical treatments or to young children with their limited attention span. As regards other portable devices for mea- suring papillae density [3,12,13], none provide the flexi- * Corresponding author. School of Women and Children’s Health, Faculty of Medicine, University of New South Wales, Level 3, Sydney Children’s bility of the digital camera. Accordingly, as an alternative Hospital, High Street, Randwick, NSW 2031, Australia. Fax: +61 2 9382 1401. method for obtaining images of taste papillae, in Part 1 here, E-mail address: [email protected] (D.G. Laing). we have investigated the use of a digital camera.

0006-8993/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2005.06.031

207 Paper 1: Shahbake et al., 2005 Appendix E

M. Shahbake et al. / Brain Research 1052 (2005) 196–201 197

Following validation of the digital camera as a suitable 2.2. Methods tool for measuring fungiform papillae density, in Part 2 of the study, the aim was to use the camera to locate the The subjects were 9 children aged 8–9 years (mean: region(s) of the anterior tongue which provide the most 8.7 years) and 7 adults aged 25–38 years (mean: 31.0 reliable location for determining papillae density and that years), from local suburbs, who were examined in a give the highest correlation with the total number of papillae University microscopy laboratory. Prior to commencing on the anterior tongue. The reason for wanting to define measurements, subjects rinsed their mouth with deionized such a region is that a number of studies have reported using water (Milli-Ro-6 Plus System, conductivity 0.9 AS). a single small region of the anterior tongue to provide an Their tongue was dried with a filter paper by the indicator of the total number of papillae in this area, experimenter, and a 6 mm diameter circular piece of however, none have demonstrated that the small region they filter paper (Whatman’s No. 1) [4] that contained a blue have used is the most reliable indicator [1,8,14]. Accord- food dye (Robert’s Brilliant Blue FCF133) was placed on ingly, by establishing such a location, we proposed that this the tip of the anterior part of the left side of the tongue would provide a reliable and rapid method for assessment of closest to the midline (Fig. 1A) for 3 s. On removal of taste function in studies of human responses to PROP and the filter paper, the tongue was again dried. In the degeneration/regeneration of taste papillae during treatment procedure used with the digital camera, to minimize head with medications and diseases which affect taste function. movement during filming, a subject supported their head by placing their arms on a table and held their head with their hands such that their chin protruded forward. The 2. Part 1 subject then protruded their tongue and held it steady with their lips (Fig. 1A). A 10 mm  3 mm wide piece 2.1. Aim of filter paper placed on the right side of the anterior tongue provided a scale to calculate the magnification of To determine if a digital camera can be used as a reliable each image (Fig. 1A). Following this, 3–5 images of the tool for measuring fungiform papillae density in humans. stained area were recorded with a Nikon Coolpix 4500

Fig. 1. (A–D) Fungiform papillae in a human tongue. (A) Tongue with midline highlighted with a white line and showing the 6 mm diameter stained area where papilla counts were conducted and the 10 mm scale. (B, C) Images of the stained area obtained with the videomicroscope and digital camera, respectively. (D) Inset in (C) viewed at the highest magnification used with the camera (Â22.5) to count papillae. Arrows indicate typical fungiform papillae.

208 Paper 1: Shahbake et al., 2005 Appendix E

198 M. Shahbake et al. / Brain Research 1052 (2005) 196–201

(4.0 megapixels) camera with a macro light attachment similar to that reported for the tongue tip using video- (Macro Cool-light SL-1). The digital images were down- microscopy (41.1) [7]. loaded to a computer (Toshiba Satellite Pro 6100) using a Thus, it was concluded that the camera can replace the USB port, and the images analyzed using a Fzoom_ videomicroscope to count papillae and is a suitable instru- option in the Adobe Photoshop 7.0 program. The ment for determining the location of a small region(s) of the magnification used in the camera was commonly Â9to anterior tongue which would be a reliable indicator of 15.5 but occasionally was up to Â24.8 when the presence fungiform papillae density. of papillae was more difficult to confirm (Fig. 1D). These higher magnifications were achieved by using the Fzoom_ option in the Adobe Photoshop program, which increased 3. Part 2 the original magnification by a factor of 1.6. In the procedure used with the videomicroscope [10], a subject 3.1. Aim was seated in a chair with their head and neck supported by a neck pillow to minimize head movement. After Determine the location on the dorsal region of the staining and drying of the tongue, a subject protruded anterior tongue which provides the most reliable measure of their tongue which was filmed for 2–8 min with a 30 s fungiform papillae density that correlates highly with the rest every minute. The average total time to obtain the total number of papillae in the region. images from a subject with the video system was 30–60 min which included microscope and head adjustments 3.2. Methods and staining and rinsing the tongue, compared to 10–15 min with the camera. The magnification used was Â30, The subjects were 30 adults aged 20–24 years (mean age the lowest possible with the videomicroscope system. 21.63 years), who were students at the University of Both imaging procedures were used at a single session to Western Sydney. ensure that precisely the same region of stained tongue Before any measurements were conducted, subjects was assessed for a subject. rinsed their mouth with deionized water (Milli-Ro-6 Plus System, conductivity 0.9 AS) and their tongue was dried 2.3. Identification of fungiform papillae with a filter paper. Following this, a quarter of a piece of filter paper (Whatman’s No. 1; area = 23.76 cm2) that Fungiform papillae were mostly mushroom-shaped and contained a blue food dye (Robert’s Brilliant Blue FCF133) elevated structures [5]. Some papillae were flat and short was placed on the left side of the anterior tongue for about with little elevation, others were double papillae [9] (Figs. 3 s. On removal of the filter paper, the tongue was dried and 1B–D). In addition, fungiform papillae were readily the stained area (Fig. 2A) photographed using a digital distinguished from filiform papillae by their very light camera (see Part 1). staining with the food dye compared to the latter papillae Prior to image analysis, a computerized grid was super- which stained dark [8]. With each method, the best image imposed on top of the blue tongue image using the Fgrid_ from an individual was used for counting papillae. option in the Fview_ menu in the Adobe Photoshop 7.0 Counting from different images collected by a specific program (Fig. 2B). The midline of the tongue, and each of method, e.g. camera, which were coded with a 3-digit the 3 most anterior centimeters of the stained area, was then number, was conducted randomly across subjects with the marked using the horizontal and vertical Fguide_ option in identity of the subject unknown to the analyst. Grouping the Fview_ menu, resulting in 3 1-cm horizontal bands and of data for statistical analysis was only done when all 12 Fquarters_ (Figs. 2B and D). Each of the bands and images from a filming method had been processed and Fquarters_ was then analyzed using a Fzoom_ option in the counting completed. Fview_ menu. Using this procedure, all the fungiform papillae in the anterior 3 cm of the stained region were 2.4. Results and discussion counted. The magnification used for the image analysis ranged between Â10.24 and Â18.43 (mean = Â14.12). The mean fungiform papilla densities and standard errors for the 6 mm diameter stained section of the tongue for the 3.3. Statistical analyses 16 subjects was 12.94 T 0.93 and 12.00 T 0.91, using the digital camera and videomicroscope, respectively. An First, for each subject, the following counts were made: independent t test indicated that the two means were not total number of fungiform papillae in the 3 cm anterior significantly different (t = 1.29, P > 0.05. The digital stained area and the number of fungiform papillae in each camera, therefore, produced similar high quality images and centimeter band of the stained area. The analysis conducted results to those obtained with the videomicroscope. Impor- was based on the first 3 cm of the tongue because a number tantly, conversion of the two values of papillae density to of subjects complained of discomfort during the measure- papillae/cm2 (45.7 and 42.4) shows that they are very ments of areas posterior to this distance.

209 Paper 1: Shahbake et al., 2005 Appendix E

M. Shahbake et al. / Brain Research 1052 (2005) 196–201 199

Fig. 2. (A–D) Analysis of fungiform papillae density. (A) Shows a tongue stained with the blue dye on the left side and a 10 mm paper strip for calibrationson the right unstained side of the tongue. (B) Image of tongue with superimposed computer-produced grid showing the three 1 cm bands and 12 Fquarters_. (C) Image of tongue showing the locations of the two 6 mm diameter regions used for counting papillae. (D) Diagram of panel B showing labeling of the 1 cm bands and the 12 Fquarters_.

A stepwise multiple regression analysis (SPSS vs. been used for measures of papillae density (e.g. [4]), an 12.0.1) was performed using the total number of fungiform image of the circle was superimposed on two regions of the papillae in the first 3 cm of the stained area as the 3 cm stained area using the imaging program (Fig. 2C). The dependent variable and the number in each centimeter band first region was at the tip closest to the midline on the left (Fig. 2B) as the predictors. This was performed to side of the tongue (Fig. 2C, Fquarters_ 1 and 3). This determine which centimeter band was the best predictor particular region has been used extensively in previous of overall papillae density. As a secondary analysis, each studies because of the high density of papillae present centimeter band was divided into four Fquarters_, thus [4,11,12]. The second region was sited in Fquarters_ 5 and 7) subdividing the stained area into twelve Fquarters_ (Figs. 2B (Fig. 2C) and was selected because they produced the and D). The same statistical analysis was then used in a secondary analysis where the number of papillae in each Fquarter_ was compared with that of the total number of Table 1 papillae in the anterior 3 cm stained area to determine which Summary of statistical analyses F _ 2 2 Fquarter_ showed the best correlation. Division into Quarter RRP Area (cm ) Fquarters_ was conducted because each centimeter band is 1 0.621 0.386 <0.001 0.406 a large area in which to count papillae, especially if this 2 0.476 0.226 <0.01 0.379 3 0.497 0.247 0.005 0.406 method is to be used with a large group of subjects. Thus, 4 0.133 0.018 >0.05 0.409 the centimeter bands were divided into smaller areas to 5 0.707 0.500 <0.001 0.527 determine whether the latter produce as good a correlation 6 0.591 0.349 0.001 0.529 as the larger 1 cm bands. 7 0.523 0.274 <0.05 0.529 Finally, earlier studies which incorporated measures of 8 0.503 0.253 0.005 0.542 9 0.328 0.108 >0.05 0.588 fungiform papillae density [4,12,13] used small stained 10 0.324 0.105 >0.05 0.592 circles on the tongue. Consequently, the present data were 11 0.494 0.244 <0.01 0.588 further analyzed using the latter statistical procedure to 12 0.485 0.235 <0.01 0.601 check the reliability of previous methods. Since a 6 mm 6 mm (1st cm) 0.501 0.251 0.005 0.283 diameter circle of filter paper with blue dye has commonly 6 mm (2nd cm) 0.468 0.219 <0.01 0.283

210 Paper 1: Shahbake et al., 2005 Appendix E

200 M. Shahbake et al. / Brain Research 1052 (2005) 196–201 highest correlations in the initial analyses of Fquarters_. Finally, the numbers of fungiform papillae and their Table 1 shows that the 6 mm circle was smaller in area than location on the surface of anterior tongue were found to any of the Fquarters_. differ among subjects as was reported earlier [7]. Despite this, as indicated above, good correlations were found 3.4. Results and discussion between the number of papillae in small areas of the tongue and the total number of papillae within the first 3 cm from The mean number of fungiform papillae in the 3 cm the tip of the tongue. stained area was 156.00 T 5.86 (SE) with a range of 106 to 229 papillae, which compares well with results from a previous videomicroscopy study which reported a mean of 4. General discussion 166.7 [7]. As regards centimeter bands 1, 2, and 3, their mean papillae numbers T SE were 91.30 T 3.88, 48.89 T The study produced two major outcomes. First, it 3.01, and 15.83 T 1, respectively. demonstrated that images of the anterior human tongue The regression performed between total papilla numbers from a 4 megapixel digital camera were of sufficient in the first 3 cm of the anterior tongue and papilla numbers quality to obtain similar fungiform papillae densities to in the first, second and third centimeter bands showed that those produced by the usual method of videomicroscopy. the second centimeter band had the highest correlation with Secondly, it showed that reliable measures of fungiform 2 the total number of papillae [R = 0.620, F(1,28) = 45.69, P < papillae density on the anterior tongue can be obtained 0.001]. The addition of the first centimeter band signifi- from relatively small areas of this region. cantly increased the variance explained in the model from The importance of demonstrating that the camera can be 2 62.0% to 96.3% [R = 0.963, F(1,27) = 246.98, P < 0.001]. used instead of a videomicroscope in measures of papillae Although the addition of the third centimeter band was also counts is that (a) significant time-saving will substantially 2 significant [R = 0.991, F(1,26) = 124.67, P < 0.001], the reduce labor costs of taste studies involving anatomical impact of it on the variance explained in the model was assessment of the tongue. (b) The size and portability of only minor compared to the first centimeter band (R2 the camera allow filming of the tongue of subjects outside change = 2.8%). the laboratory in clinics, at the bedside of patients or in Regression analyses of the Fquarters_ (Table 1) showed schools. (c) The camera costs substantially less than the that the first Fquarter_ in the second centimeter band videomicroscope. (Fquarter_ 5) had the highest correlation with the total The finding that an indication of the papillae density of 2 number of papillae [R = 0.500, F(1,28) = 28.035, P < the anterior tongue can be obtained from relatively small 0.001]. The variance explained in the model, however, areas within much of this region has two important increased significantly from 50 to 79.1% [R2 = 0.791, implications. First, previously, various workers [3,7,12,13] F(1,27) = 50.96, P < 0.001] when Fquarter_ 1 was included have arbitrarily chosen an area near the tip of the tongue as in the analysis. Even though these Fquarters_ showed a an indicator of overall fungiform papillae density or of the strong correlation with the total number of fungiform total number of fungiform papillae. However, no validation papillae, seven of the other Fquarters_ also showed signi- of any of the regions used has been reported. The present ficant relationships ( P < 0.01) (Table 1). Only Fquarters_ 4, data, therefore, provide future studies with a clear insight to 9, and 10 showed weaker non-significant correlations locations that can be used with surety. In this regard, it with the total number of papillae ( P > 0.05). Overall, confirmed that the 6 mm diameter area used by a number of the results indicated that counts of fungiform papillae in workers to assess papillae density also provides a reliable almost any section of the first 3 cm from the tip of the measure of density on the anterior tongue (Table 1). tongue (except Fquarters_ 4, 9, and 10) will produce a Secondly, the significant correlations between many of the reliable indicator of the fungiform papillae density of the areas assessed and total papillae number or density were a anterior tongue. surprise given the wide variation in the density of papillae As regards the relationship of the densities of the across the anterior tongue. Nevertheless, the results indicate papillae in the two regions involving the 6 mm diameter that, provided the same (equivalent) area is used within a circle to the overall number of papillae in the 3 cm group of individuals and the area is one or a combination of stained area, the analyses showed that there was a areas that showed significant correlations here, a reliable significant correlation between both regions and the comparison of total papillae number or overall papillae overall fungiform papillae density [Position 1: R2 = density between individuals and between groups can be 2 0.251, F(1,28) = 9.388, P = 0.005; Position 2: R = obtained. 0.219, F(1,27) = 7.834, P = 0.009]. Importantly, these The use of fungiform papillae density as a tool for results indicate that the locations used in previous research assessing taste function is not advocated here as the methods for measuring local papillae density are reliable yardstick for comparing absolute taste function in one and can be used as an indicator of overall papillae counts individual with that of another. Other studies, for example, on the anterior tongue. have shown that there is a substantial variation of papillae

211 Paper 1: Shahbake et al., 2005 Appendix E

M. Shahbake et al. / Brain Research 1052 (2005) 196–201 201 numbers and density across individuals [6] and that the [2] A.D. Conger, Loss and recovery of taste acuity in patients irradiated to relationship between absolute sensitivity or perceived the oral cavity, Radiat. Res. 53 (1973) 338–347. [3] G.K. Essick, A. Chopra, S. Guest, F. McGlone, Lingual tactile acuity, intensity and the two anatomical measures is at best a taste perception, and the density and diameter of fungiform papillae in qualitative rather than quantitative one [1,7,14]. Thus, the female subjects, Physiol. Behav. 80 (2003) 289–302. most common current uses of these latter measures are to [4] K. Fast, V.B. Duffy, L.M. Bartoshuk, New psychophysical insight in assist in the classification of individuals as supertasters and evaluating generic variation in taste, in: C. Rouby, B.M. Schaal, D. non-tasters of PROP, sensitivity to which has been shown to Dubois, R. Gervais, A. Holley (Eds.), Olfaction, Taste and Cognition, Cambridge Univ. Press, New York, 2002, pp. 391–407. have a genetic basis [1], and allow changes within an [5] I.J. Miller, Anatomy of the peripheral taste system, in: R.L. Doty individual or group of individuals to be monitored and to be (Ed.), Handbook of Olfaction and Gustation, Marcel Dekker, New correlated with psychophysical data. Typical applications York, 1995, pp. 521–547. include monitoring the progress of taste loss and recovery [6] I.J. Miller, L.M. Bartoshuk, Taste perception, taste bud distribution, following lingual nerve section [15], during chemo- or and spatial relationships, in: T.V. Getchell, L.M. Bartoshuk, R.L. Doty, J.B.J. Snow (Eds.), Smell and Taste in Health and Disease, Raven radiotherapy in cancer treatment [2], and the administration Press, New York, 1991, pp. 205–233. of medications that affect taste [9]. [7] I.J. Miller, F.E. Reedy, Quantification of fungiform papillae and taste In conclusion, the digital camera appears to be a valuable pores in living human subjects, Chem. Senses 15 (1990) 281–294. tool whose characteristics will undoubtedly lead to its use in [8] I.J. Miller, F.E. Reedy, Variation in human taste bud density and taste many studies of taste function where the aim is either to intensity perception, Physiol. Behav. 47 (1990) 1213–1219. [9] S.S. Schiffman, B.G. Graham, M.S. Suggs, E.A. Sattely-Miller, Effect correlate objective anatomical data with subjective psycho- of psychotropic drugs on taste responses in young and elderly persons, physical measures or to estimate taste function using a non- Ann. N. Y. Acad. Sci. 855 (1998) 732–737. invasive and relatively simple and rapid technique. [10] C. Segovia, I. Hutchinson, D.G. Laing, A.L. Jinks, A quantitative study of fungiform papillae and taste pore density in adults and children, Dev. Brain Res. 138 (2002) 135–146. Acknowledgments [11] S.I. Sollars, I.L. Bernstein, Neonatal chorda tympani transection permanently disrupts fungiform taste bud and papilla structure in the rat, Physiol. Behav. 69 (2000) 439–444. The authors wish to thank the participants for their [12] B.J. Tepper, R.J. Nurse, Fat perception is related to PROP taster status, assistance. MS was supported by a Postgraduate Scholarship Physiol. Behav. 61 (1997) 949–954. from the Centre For Advanced Food Research, and the [13] C.A. Yackinous, J.X. Guinard, Relation between PROP taster status, research was approved by the University Human Research taste anatomy and dietary intake measures for young men and women, Ethics Committee (Approval No. HEC02/188). Appetite 38 (2002) 201–209. [14] J.R. Zuniga, S.H. Davis, R.A. Englehardt, I.J. Miller, S.S. Schiff- man, C. Phillips, Taste performance on the anterior human tongue varies with fungiform taste bud density, Chem. Senses 18 (1993) References 449–460. [15] J.R. Zuniga, N. Chen, C.L. Phillips, Chemosensory and somatosen- [1] L.M. Bartoshuk, V.B. Duffy, I.J. Miller, PTC/PROP tasting: anatomy, sory regeneration after lingual nerve repair in humans, J. Oral psychophysics and sex effects, Physiol. Behav. 56 (1994) 1165–1171. Maxillofac. Surg. 55 (1997) 2–13.

212 Paper 2: Correa et al., 2013 Appendix E Chem. Senses 38: 519–527, 2013 doi:10.1093/chemse/bjt022 Advance Access publication May 24, 2013

Changes in Fungiform Papillae Density During Development in Humans

Maryam Correa1, Ian Hutchinson1, David G. Laing1,2, and Anthony L. Jinks3 1Children’s Food Research and Education Unit, Centre for Advanced Food Research, College of Science and Technology and Environment, University of Western Sydney, Locked Bag 1797, Penrith South, NSW 1797, Australia, 2School of Women’s and Children’s Health, Faculty of Medicine, University of New South Wales, Level 3, Sydney Children’s Hospital, High Street, Randwick, NSW 2031, Australia and 3School of Psychology and Social Sciences, University of Western Sydney, Locked Bag 1797, Penrith South, NSW 1797, Australia Downloaded from

Correspondence to be sent to: David G. Laing, School of Women’s and Children’s Health, Faculty of Medicine, University of New South Wales, Level 3, Sydney Children’s Hospital, High Street, Randwick, NSW 2031, Australia. e-mail: [email protected]

Accepted April 24, 2013 http://chemse.oxfordjournals.org/

Abstract The anterior region of the human tongue ceases to grow by 8–10 years of age and the posterior region at 15–16 years. This study was conducted with 30 adults and 85 children (7–12 year olds) to determine whether the cessation of growth in the anterior tongue coincides with the stabilization of the number and distribution of fungiform papillae (FP) on this region of the tongue. This is important for understanding when the human sense of taste becomes adult in function. This study also

aimed to determine whether a small subpopulation of papillae could be used to predict the total number of papillae. FP were at University of Sydney Library on October 29, 2015 photographed and analyzed using a digital camera. The results indicated that the number of papillae stabilized at 9–10 years of age, whereas the distribution and growth of papillae stabilized at 11–12 years of age. One subpopulation of papillae pre- dicted the density of papillae on the whole anterior tongue of 7–10 year olds, whereas another was the best predictor for the older children and adults. Overall, the population, size, and distribution of FP stabilized by 11–12 years of age, which is very close to the age that cessation of growth of the anterior tongue occurs.

Key words: adults and children, gustatory development, papillae density predictors, tongue

Introduction anterior tongue with the highest densities being recorded Fungiform papillae (FP) contain taste-sensing receptor cells, near the tip (Miller 1986; Stein et al. 1994). However, there which are accessible by tastants via pores in taste buds. They can be wide differences between the distributions of papillae are characteristically found on the tongue of mammals and of individual subjects with some having very high densities are visible to the eye on the dorsal surface. In humans, the near the tongue tip, whereas others exhibit more even majority of FP are located on the most anterior area (Miller distributions of papillae across the anterior tongue (Miller and Reedy 1990a), which extends to about 2 cm from the and Reedy 1990a). Furthermore, because 99% of papillae in tongue tip (Temple et al. 2002). In contrast, they are only humans contain at least 1 taste bud (Segovia et al. 2002), sparsely distributed over the posterior area. Importantly, the papillae density is related to taste sensitivity with subjects anterior tongue ceases to grow at 8–10 years of age, whereas who have higher numbers of papillae being more sensitive the posterior region continues to grow until 15–16 years of to taste stimuli (Smith 1971; Stein et al. 1994; Delwiche age (Temple et al. 2002). A critical finding in the latter study et al. 2001). Thus, because papillae density varies across the was that the area of the anterior tongue in 8–10 year olds tongue, sensitivity across the tongue also varies. A question and adults is the same. Accordingly, this equivalence in size that has not been resolved is whether the densities of papillae allows direct comparison of papillae density in any part of across the anterior tongue in children and adults are similar. the anterior tongue of adults and children of this age. In If differences in papillae distributions occur between adults and children, FP density commonly varies across the adults and children, they are also likely to affect

© The Author 2013. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: [email protected] 213 Paper 2: Correa et al., 2013 Appendix E 520 M. Correa et al. estimates of the total number of FP in the anterior adults found here are better than those used by others tongue. Previous estimates of the total number of papil- with adults. lae in adults have mostly been obtained by measuring the number of papillae in a randomly selected single small region and using this as an indicator of density on the Materials and methods whole anterior tongue, for example, Tepper and Nurse 1997; Fast et al. 2002; Yackinous and Guinard 2002. General However, in none of these studies has the region chosen There were 115 participants (69 females and 46 males) com- been shown to be the best for predicting overall papil- prising 30 adults aged 20–24 years (21.63 ± 0.95 [standard lae density. More recently, Shahbake et al. (2005) used error {SE}] years; 21 females and 9 males) and 85 children of a digital camera and the same software as described in which 29 were 7–8 years old (7.78 ± 0.47 years; 17 females and this study to show that 7/12 small regions across the 12 males), 23 were 9–10 years old (9.68 ± 0.51 years; 15 females anterior tongue of adults were better predictors of the and 8 males) and 33 were 11–12 years old (11.60 ± 0.49 years; total papillae numbers than the 2 areas previously used. 16 females and 17 males). The adults were university students Downloaded from Establishing the age at which papillae densities in regions who responded to poster advertisements around the cam- on the anterior tongue stabilize would provide evidence pus and were examined in a university microscopy labora- that full maturity has been achieved. With the discovery tory. The children were from local public schools and were that growth of the anterior tongue ceases at 8–10 years of examined in a quiet room at their school. Written consent age, this can now be achieved. Thus, as indicated above, it was obtained from each adult and a parent of each child once http://chemse.oxfordjournals.org/ is now possible to compare equivalent tongue regions of they had read an information sheet describing the purpose of adults and children as regards their regional papillae den- the study and details of how measurements would be con- sities. Accordingly, one of the major goal of this study is ducted. Only subjects with obvious flu/colds/middle ear infec- to count the total number of FP in the anterior tongue tions, or oral problems such as recent withdrawal of teeth, of adults and children in equivalent small regions. This which could have damaged the chorda tympani that inner- should provide the evidence necessary to demonstrate vates FP and possibly caused loss of papillae, were excluded. at what age stabilization of papillae density occurs. The Subjects within each age group were not fully matched for advantage of being able to measure papillae density in a gender for several reasons; 1) prior to the study, Temple small area to predict overall papillae numbers/densities is

(1999) had found that the number of papillae in the anterior at University of Sydney Library on October 29, 2015 that it would provide a rapid time-saving procedure for tongue of matched female and male 8 year olds was the same; monitoring taste loss and recovery in clinical conditions 2) Temple et al. (2002) found no gender differences as regards that include lingual nerve section (Zuniga et al. 1997), the size/area of the anterior tongue of age-matched groups function during chemo- and radiotherapy in cancer treat- of female and male 8–10 year olds, or in older children and ment (Conger 1973), and administration of medications adults. All participants received compensation (movie tickets that affect taste (Schiffman et al. 1998). Furthermore, and confectionary) at the conclusion of their participation. because papillae densities reflect taste sensitivity, the The research was approved by the University of Western ability to use a small area of the tongue to gain an insight Sydney Human Research Ethics Committee (Approval No. to the ability of a human to perceive tastes will be useful HEC02/188) and conducted in accordance with the ethical in studies where the aim is to correlate objective anatomi- standards laid down in the 1984 Declaration of Helsinki. cal data with subjective psychophysical measures such as Prior to commencing the photographing and counting of in classification of supertasters, tasters, and non-tasters papillae, subjects rinsed their mouth with deionized water in studies using propthiouracil (PROP). Accordingly, a (Milli-Ro- Plus System, conductivity 0.9 µS), and their tongue goal of this study is to 1) determine whether the regions was dried with a filter paper by the experimenter. A piece of that best predict overall papillae density in adults are filter paper (Whatman’s No. 1) that contained a blue food the same in children, and at what age this occurs and dye (Robert’s Brilliant Blue FCF133) was placed on the 2) determine whether the best predictors for children are left side of the anterior tongue for 3 s (Figure 1). The area better than the 2 regions used previously by others with stained was the most anterior 3–4 cm of the dorsal tongue. adults. Such knowledge would also assist in defining the This length was used because it adequately covered more age at which taste development on the tongue ceases. This than 2 cm “demarcation line,” which defined the anterior and study consists of 3 parts. In Part 1, we describe when the posterior regions of the tongue in both adults and children number and distribution of papillae stabilize and become (Temple et al. 2002). On removal of the filter paper, the the same in children as in adults; in Part 2, we determine tongue was dried with a clean filter paper. Head movement whether a small region(s) of the anterior tongue can be during photography of the tongue was minimized by the used to reliably predict the papillae density on the whole participant placing their arms on a table and holding their anterior tongue of adults and children; and in Part 3, we head with their hands such that their chin protruded forward. determine if the best predictor regions for children and The participants extended their tongue and held it steady with

214 Paper 2: Correa et al., 2013 Appendix E Changes in Fungiform Papillae Density 521 their lips while 3–7 images of the 2-cm anterior stained area were “marked” using the horizontal and vertical “guide” were captured with the camera. Images were taken at different option in the “view” menu of Adobe Photoshop resulting angles to ensure that all of the 2-cm stained area was visible in two 1-cm horizontal bands (Figure 1). Because the area in the images captured. A 10 × 3 mm wide piece of filter paper of the anterior tongue was the same in adults and children placed on the right side of the anterior tongue provided a (Temple et al. 2002), the software adjusted the 2 bands to be scale to calculate the magnification of each image (Figure 1). equivalent for all subjects. In essence, the software provided Images were recorded with a Nikon Coolpix 4500 (4.0 a “virtual template” that adjusted the size of the bands to be megapixels) camera with a macro light attachment (Macro equivalent across subjects. The number of FP in each 1-cm Cool-light SL-1). The digital images were downloaded to a region were identified and counted using the “zoom” option computer (Toshiba Satellite Pro 6100) and analyzed using in the “view” menu of Adobe Photoshop. All images were a “zoom” option in the Adobe Photoshop 7.0 program coded with a 4-digit number and randomized before analysis (Shahbake et al. 2005). The best image of the 2-cm anterior so that the identity of the participants and their age groups tongue from an individual was the image that was used for were unknown to the experimenter. counting papillae in each part of the study. Accordingly, the Downloaded from data collected in the 3 parts were from the same subjects and Results and discussion images were recorded at this one and only photography ses- sion. The magnification used for the image analysis ranged For data analysis, the participants were divided into 4 age between ×14.12 and ×23.73 (mean = ×18.43). groups: 7–8, 9–10, 11–12 years, and adults, and for each par- FP were identified as mostly mushroom-shaped and ele- ticipant, the number of FP in each 1-cm region and total http://chemse.oxfordjournals.org/ vated structures (Miller 1995; Segovia et al. 2002; Shahbake number of papillae in the anterior tongue were determined et al. 2005). Some papillae were flat and short with little (Table 1). elevation, others were double papillae. In addition, FP were readily distinguished from filiform papillae by their very light staining with the food dye compared with the latter papillae, Total number of FP which stained dark (Miller and Reedy 1990b). A 2 × 4 (gender × group) analysis of variance (ANOVA) was performed amongst all age groups to determine if there were Part 1 significant differences in the total number of FP in the ante-

rior tongue. This indicated that there was a significant differ- at University of Sydney Library on October 29, 2015 Aims ence between age groups (F[3,107] = 3.238, P = 0.025; Table 1). Determine 1) whether the number of FP in the whole ante- Post-hoc Tukey Least Squares Difference-paired compari- rior tongue differs for adults and children and 2) whether FP son tests (P < 0.05) indicated that 7–8 year olds had signifi- density in two 1 cm wide regions of the anterior tongue is cantly more papillae than adults (P = 0.01), and there were different and change throughout childhood. no significant differences between any other groups. This is an important finding because it indicates the total number of FP in humans stabilized during childhood at 9–10 years of Materials and methods age. The ANOVA also indicated that there was no significant

Prior to image analysis, the midline of the tongue and gender differences (F[1,107] = 0.033, P = 0.855; Table 1) and no each of the 2 most anterior centimeters of the stained area gender × group interactions (F[7,107] = 0.974, P = 0.408).

Figure 1 Analysis of FP density. (A) Image of the tongue showing the most anterior stained area on the left side divided into two 1-cm intervals as in Part 1 and 8 smaller areas as in Part 2. On the right side is the 1-cm scale. (B) A diagram of the tongue showing the numbering sequence of the 8 small areas.

215 Paper 2: Correa et al., 2013 Appendix E 522 M. Correa et al.

Table 1 Mean number of FP in the anterior tongue and in two 1-cm regions for different age groups Group Total FP FP range First cm FP Second cm FP Total FP M/F ± (SE) counts ± (SE) counts ± (SE) counts ± (SE) Male Female

7–8 years 162.24 ± 3.74 116–195 109.93 ± 2.63 52.31 ± 2.90 159.08 ± 6.41 164.47 ± 4.59 9–10 years 146.09 ± 4.69 75–186 93.87 ± 3.09 52.22 ± 2.43 140.07 ± 5.21 157.38 ± 8.33 11–12 years 150.79 ± 5.27 85–217 93.39 ± 4.24 57.40 ± 2.81 149.71 ± 6.27 151.94 ± 8.81 Adults 140.20 ± 5.50 93–222 91.30 ± 4.07 48.90 ± 3.01 134.11 ± 8.25 141.95 ± 7.11

Regional analysis anterior tongue region. The patterns were recorded by “mark- Two 2 × 4 (gender × group) between-participants ANOVAs ing” each fungiform papilla with a dot and the tongue mid- were performed to determine if there were significant differ- line and 2-cm region were “marked” to enhance consistency ences 1) in the FP density of the age groups within the first and of regions between subjects. The “markings” were superim- Downloaded from second 1-cm regions of the anterior tongue and 2) between posed on each image. One pattern had clusters of closely females and males of all groups within these regions. packed FP at the tip of the tongue with a sharply decreasing The ANOVA indicated that for the most anterior 1-cm number toward the posterior tongue. A second pattern had region (tongue tip), the number of papillae differed between no clusters, and the papillae were distributed approximately

evenly over the anterior 2 cm of the tongue. No other type http://chemse.oxfordjournals.org/ the groups (F[3,107] = 5.282, P = 0.002). Post-hoc Tukey tests (P = 0.05) showed that 7–8 year olds had significantly more of distribution was discerned and there appeared to be no papillae than the 3 older groups and the latter groups had gender-related differences in the patterns of papillae distribu- similar numbers of papillae. As regards gender, there were tion. The finding of variations in the distributions of FP is no significant differences between females and males within not new and has been reported for adults (Miller and Reedy each age group (F = 0.021, P = 0.885). Interestingly, there 1990a) and both adults and 8 year olds (Stein et al. 1994). [1,107] In summary, only the 7–8-year-old group had more papil- was a significant gender × group interaction F( [7,107] = 2.993, P = 0.007). Accordingly, an interaction graph was constructed lae than adults, and this difference was due to the youngest to investigate where these differences appear. This showed that group having more papillae in the 1-cm region nearest the with males, FP density decreased from 7–8 year olds to adults, tongue tip. The shape and size of papillae were smaller and at University of Sydney Library on October 29, 2015 whereas with females, there was a density decrease from 7–8 more regular in 7–10 year olds than in 11–12 year olds and to 9–10 years, where it remained stable through to adulthood. adults. From these data, it can be concluded that the num- As regards the second 1-cm region, a 2 × 4 (gender × group) ber of papillae stabilizes by 9–10 years of age, whereas the ANOVA indicated that there were no significant differences growth and shape of papillae stabilizes by 11–12 years. In none of the measures was there a gender difference. between the papillae numbers of the groups (F[3,107] = 1.288, P = 0.282), or between females and males (F[1,107] = 0.249, P = 0.619), and there were no significant interactions Part 2

(F[7,107] = 0.804, P = 0.586). In brief, because 7–8 year olds had more papillae on the most anterior 1-cm region than the Aims other age groups but had similar numbers to the other groups Determine 1) whether there is a subpopulation(s) of FP on in the second more posterior 1-cm region, it can be concluded the anterior tongue whose density is a reliable predictor of that papillae numbers stabilize at 9–10 years of age. the total number of papillae and 2) whether the location of Another feature of the papillae was a variation in their the subpopulation (s) is dependent on age. shape with age. In the youngest group, papillae were iden- These questions were of interest for several reasons. First, tified as small and round structures. In 9–10 year olds, the in Part1, papillae were counted in 2 relatively large regions papillae were noticeably larger in size, whereas in 11–12 year of the anterior tongue and the time required to count the olds and adults, many papillae were irregular in shape, papillae in all the subjects was considerable. Accordingly, which was similar to previous findings from this laboratory it was important to determine if a smaller regional (Segovia et al. 2002). The increase in size with age is in agree- subpopulation(s) of papillae would prove to be a suitable ment with the finding that adult papillae have significantly predictor of overall papillae density. The data should also larger diameters than those of 8–9 year olds (Segovia et al. show when the papillae density across the anterior tongue 2002). Thus, although the number of papillae stabilizes by stabilizes signaling maturation of the peripheral gustatory 9–10 years of age, changes in the shape and size of the papil- system. To achieve a more fine-grained analysis, the two lae continue until 11–12 years. 1-cm anterior regions described in Part 1 were divided Adults and children were also observed to have 2 qualita- into 8 smaller regions (Figure 1). Also, because it had been tively different distribution patterns of FP across the 2-cm shown that for adults the best predictor of total papillae

216 Paper 2: Correa et al., 2013 Appendix E Changes in Fungiform Papillae Density 523 numbers was a subpopulation located in the second 1-cm Table 3 Predictor models of total FP on the anterior tongue region (Shahbake et al. 2005), another goal was to expand Group (year) Model R2 this finding to determine if the same result was obtained for 7–8 1: y = (2.281A + 60.850) 0.44 children of different ages. This was of particular interest 1 because in Part 1, it had been shown that papillae numbers 2: y = (2.259A1 + 2.048A5 + 31.878) 0.56 varied with age with changes being observed in the most 3: y = (2.009A1 + 2.990A5 + 1.904A7 + 7.429) 0.69 anterior 1-cm region for 7–8 year olds. 9–10 1: y = (1.701A1 + 84.563) 0.45

2: y = (1.779A1 + 2.422A5 + 50.787) 0.62 Materials and methods 3: y = (1.806A1 + 1.916A5 + 2.210A7 + 9.959) 0.71 Each of the two 1-cm regions on the anterior tongue was 11–12 1: y = (3.067A + 100.215) 0.52 divided into 4 smaller areas using the “grid” option in the 5 “view” menu of Adobe Photoshop (Figure 1). These 8 areas 2: y = (2.522A5 + 2.516A4 + 58.28) 0.72 were numbered in a manner where Areas from 1 to 4 were in 3: y = (2.123A + 2.477A + 2.141A + 25.932) 0.81

5 4 2 Downloaded from the most anterior 1-cm region, and Areas from 5 to 8 were in Adults 1: y = (3.190A + 92.569) 0.46 the second region. Papillae were counted in each of the 8 areas, 5 which were then compared with the total number of papillae 2: y = (2.195A5 + 0.865A1 + 61.313) 0.60 in the 2-cm anterior tongue region. The length and width of 3: y = (2.894A5 + 0.638A1 + 1.203A2 + 48.321) 0.69 each small area were measured based on the 1-cm scale. This allowed the area of each of the 8 areas to be calculated and the http://chemse.oxfordjournals.org/ areas in the calculations increased predictability. For example, density of papillae per cm2 to be determined (Table 2). in the case of 7–8 year olds, predictability increased from 0.44 to 0.56 when Area 5 was included, and for adults, predictabil- Results and discussion ity increased from 0.46 to 0.60 when Area 1 was included. The data were also analyzed by including data from all age groups, For each of the 4 age groups, a stepwise multiple regression and this indicated that the best overall predictor was Area 1 analysis (SPSS 12.0.1) was used to identify which Area(s) best (R2 = 0.33). Inclusion of Area 5 in the calculations increased predicted the total number of papillae in the 2-cm anterior R2 to 0.51. As expected from previous studies (Miller 1986; tongue region, and for each of the 8 areas, the best 3 predictor Miller and Reedy 1990a; Stein et al. 1994; Segovia et al. 2002; at University of Sydney Library on October 29, 2015 models were calculated (Table 3). The models were based on Shahbake et al. 2005), for all age groups, the number of papil- the equation y = bA + bA + bA + … + Constant, where b 1 2 3 lae per area decreased in the anterior to posterior direction. was the coefficient calculated from the regression analysis, A x It is concluded that the FP density of the whole anterior tongue represented the number of the Area, and the constant was cal- can be predicted from 1 or at most 2 small areas on the tongue, culated from the regression analysis. As shown in Table 3, Area 2 with different areas being better predictors depending on the 1 was the best predictor of total FP density for 7–8 (R = 0.44) 2 age of subjects. Because the best predictor area for 11–12-year- and 9–10 (R = 0.45) year olds, respectively, with Area 5 the 2 old children is the same as for adults, it can be concluded that best predictor for 11–12 year olds (R = 0.52) and adults 2 the distribution of papillae stabilizes at 11–12 years of age. (R = 0.46). Table 3 also indicated that including 1 and 2 other

Part 3 Table 2 Mean FP density (papillae/cm2 ± SE) in 8 regional subpopulations and Areas X and Y on the anterior tongue Aim Area 7–8 years 9–10 years 11–12 years Adult Determine whether the papillae density in a 6-mm diameter 1 120.95 ± 9.11 113.69 ± 9.12 115.44 ± 6.53 118.79 ± 13.33 circular area (0.28 cm2) on the anterior tongue is 1) a reliable 2 76.48 ± 8.02 79.23 ± 5.86 74.40 ± 5.86 68.30 ± 13.73 predictor of the total number of papillae on the anterior 3 64.78 ± 4.19 62.57 ± 6.52 60.29 ± 3.67 52.47 ± 4.28 tongue in adults and children, and 2) that it is a more reliable predictor than either of the 2 predictors, that is, Areas 1 and 4 64.75 ± 3.68 65.33 ± 4.82 60.44 ± 3.61 61.39 ± 8.53 5, found in Part 2. 5 32.72 ± 2.17 32.75 ± 2.80 34.40 ± 2.58 36.15 ± 6.24 This study was conducted as an advance on earlier stud- 6 38.71 ± 3.67 37.38 ± 3.02 33.12 ± 2.88 43.30 ± 10.97 ies with adults that used the papillae density of a randomly selected small circular area of the anterior tongue to esti- 7 30.36 ± 3.10 27.23 ± 3.64 26.40 ± 2.81 21.82 ± 4.75 mate density of the whole anterior tongue (Tepper and 8 32.35 ± 2.46 34.16 ± 3.40 23.40 ± 2.33 18.44 ± 2.46 Nurse 1997; Fast et al. 2002; Yackinous and Guinard 2002). X 139.05 ± 4.32 138.11 ± 5.75 135.49 ± 5.43 133.95 ± 4.04 However, only 1 study (Shahbake et al. 2005) determined the reliability of the procedure, which was found to provide a Y 64.16 ± 2.61 62.44 ± 2.73 59.49 ± 3.05 59.43 ± 2.26 modest but statistically significant estimate.

217 Paper 2: Correa et al., 2013 Appendix E 524 M. Correa et al.

Materials and methods P = 0.643), but there was a significant gender × group inter- FP density was quantified in two 6-mm diameter circular action (F[4,139] = 1.10, P = 0.044). The interaction plot showed areas designated Area X and Area Y (Figure 2.). The figure there were similar trends for females and males with a grad- shows these 2 areas superimposed on the stained anterior ual decrease from 7–8 year olds to adults. tongue. Area X included Area 1, which was the best indica- In summary, the 2 regions selected by others to predict the tor of anterior tongue papillae density for 7–8 and 9–10 year total papillae density of the anterior tongue gave statistically olds, and a small part of Area 3. Area Y encompassed Area significant results for both adults and children. However, the correlations for Areas X (R2 = 0.24) and Y (R2 = 0.18) were 5, which was the best predictor of overall papillae density for 2 11–12 year olds and adults, and part of Area 7. Area X was lower than when the predictor was Area 1 (R = 0.33), which similar to that used to measure papillae density in adults by was the best predictor area when the data from all subjects Tepper and Nurse (1997), Yackinous and Guinard (2002), were combined. However, the most reliable correlations for Fast et al. (2002), and Shahbake et al. (2005). each age group were recorded when Area 1 was used for 7–8 (R2 = 0.44) and 9–10 (0.45) year olds and Area 5 for 11–12 (R2 = 0.52) year olds and adults (R2 =0.46). In every instance, Downloaded from Results and discussion a lower correlation was obtained for each age group when The FP density for Area X was 139.05 ± 4.32 (SE), either Area X or Y was used as the predictor area. 138.11 ± 5.72, 135.49 ± 5.43, and 133.95 ± 4.04 papillae/cm2 for 7–8, 9–10, 11–12 year olds, and adults, respectively (Table 2).

Lower densities were found in Area Y for all age groups. General discussion http://chemse.oxfordjournals.org/ These were 64.16 ± 2.61, 62.44 ± 2.73, 59.49 ± 3.05, and 59.43 ± 2.26 papillae/cm2 for 7–8 year olds to adults, respec- The study had 3 major aims, all of which were achieved. The tively. Importantly, the densities for adults found here were first was to determine at what age the number of papillae similar to those reported. Thus, the density here for Area X stabilized. The results of Part 1 indicated the number became was 133.95 papillae/cm2 compared with 99.5 and 141 by Fast constant by 9–10 years of age. The second was to determine et al. (2002) and Yackinous and Guinard (2002), respectively. whether there is a location on the tongue where the papillae Pearson’s correlation analyses showed there were signifi- density can be used to predict the overall papillae density of cant correlations between the total FP counts on the ante- the anterior tongue. Again, a clear outcome was achieved with 2 the data of Part 2 indicating that there are 2 locations, which rior tongue and both Areas X (R = 0.24, P < 0.001) and at University of Sydney Library on October 29, 2015 Y (R2 = 0.11, P < 0.001) when the data from all subjects are dependent on age. Third, it was shown in Part 3 that the were combined. Both correlations are poorer than that for method used by earlier workers for estimating overall papillae Area 1 (R2 = 0.33) when all subjects were included in the density on the anterior tongue produced reliable results, but analysis. In addition, a 2 × 4 (gender × group) ANOVA indi- these were less reliable than the 2 subpopulations of papillae cated there was no significant difference in papillae density that were found here to be the best for particular age groups. between groups in Area X (F = 0.783, P = 0.860) and This study is the first to demonstrate that the FP density [1, 107] on the human tongue reaches adult numbers by mid-child- there were no significant gender differences F( [1,107] = 0.783, P = 0.378) within the groups or gender × group interactions. hood. The results are in agreement with 2 earlier studies of As regards Area Y, papillae densities were not different papillae numbers that had shown papillae density decreased from childhood to adulthood. One study found that density across groups (F[3,107] = 0.594, P = 0.620 and there were no significant gender differences within groups F( = 0.216, decreased from 0–10 year olds to adults aged 50–60 years [1, 107] (Moses et al. 1967) and the other study showed that 8–9 year olds had significantly higher papillae densities than adults (Segovia et al. 2002). However, neither indicated at what age papillae density stabilized. Importantly, the most strik- ing feature in both studies was that the number of papillae decreased from childhood to adulthood, as was reported for circumvallate papillae (Arey et al. 1935). The mean number of FP found in the anterior tongue of adults in this study was 140.20 ± 5.50 (SE) and is simi- lar to the number reported by Miller and Reedy (1990a) using video microscopy. Comparison of the adult papil- lae number (140.20) with the number found for 7–8 year olds (162.24 ± 3.74) indicates that the children had ~13.6% more papillae, which is a statistically significant difference. Analysis of the papillae numbers in each of the two 1-cm Figure 2 Image of tongue showing the locations of Areas X and Y. regions indicated the difference was primarily due to the

218 Paper 2: Correa et al., 2013 Appendix E Changes in Fungiform Papillae Density 525 greater number of papillae in the most anterior region of densities than indicated by the R2 values for Areas 1 and 5 7–8-year-old tongues (Table 1). The finding that there were for particular age groups. The lower predictability of Area no changes in the numbers of papillae in the first 1-cm X appears to be due to the inclusion of part of Area 3 in region in children aged above 7–8 years and no changes in the papillae counts, which was not one of the Areas found the second 1-cm region with any of the age groups indicates to be a predictor for any of the age groups in the 3 models that stabilization of papillae numbers in the anterior tongue described in Table 3. Similarly, the R2 values for Area Y were had occurred by 9–10 years of age. Importantly, stabiliza- lower than found for Area 5 for 11–12 year olds (0.52 vs. tion of papillae numbers by 9–10 years of age coincides with 0.28) and adults (0.46 vs. 0.13) because part of Area 7 was completion of the growth of the “taste-sensitive” area of the included in the papillae counts, and this area was not found anterior tongue (Temple et al. 2002). to be a significant predictor for these age groups. In brief, the Another feature of the papillae that indicated changes in results indicated Areas 1 and 5 were better predictors than papillae density occurs until mid-childhood was their shape Areas X and Y for particular age groups. and size. As indicated in Part 1, papillae of 7–8 year olds were In none of the above measures of papillae density, or the identified as small round structures; in 9–10 year olds, they best predictor regions, was there a significant gender effect. Downloaded from were larger in size, whereas in 11–12 year olds and adults, Absence of a gender effect has been reported in a number they were irregular in shape. These results are in agreement of studies on tongue growth (Temple et al. 2002), taste bud with the report that 8–9 year olds have rounder and smaller density (Miller 1988; Zuniga et al. 1993), and taste buds per papillae than adults, and adults have papillae that differ papilla (Arvidson 1979) and appears to be a robust finding in shape more than in children (Segovia et al. 2002). The as regards anatomical features of the human peripheral gus- http://chemse.oxfordjournals.org/ changes in size and shape also suggest that development of tatory system. the structure of papillae is complete at 11–12 years of age. As mentioned above, an important finding of this study was The imaging procedure described in Part B that was used that taste-sensing papillae on the anterior tongue decrease to accurately divide the anterior tongue into 8 regions, which in number during mid-childhood and plateau at about were equivalent in size across subjects, produced data that 9–10 years of age. This loss of papillae, which would include for the first time provided a reliable method for predict- loss of taste buds and cells within the papillae, parallels other ing the papillae density of the whole anterior tongue from developmental phenomena in mammalian gustation and the density of a small regional subpopulation of papillae. other sensory systems. For example, 8–9 year olds have a Importantly, the data indicated that the location of the greater density of taste pores than adults (Segovia et al. 2002), at University of Sydney Library on October 29, 2015 regional subpopulation that best predicted overall papillae and the number of taste buds in FP in sheep innervated by density was dependent on age. Thus, the density of papil- a single chorda tympani neuron increases in perinatal sheep lae in Area 1 at the tongue tip was the best predictor of the as compared with fetal sheep and then decreases through density of papillae in the whole anterior tongue for children postnatal development (Mistretta et al. 1988; Nagai et al. aged 7–8 and 9–10 years, whereas Area 5 in the second 1-cm 1988). In olfaction, there is a large loss of granule cells, the region was the best predictor for 11–12 year olds and adults major interneurons in the mouse olfactory bulb, during the (Table 3). Given that a reduction in the number of papil- first 2 postnatal weeks, after which the number of granule cells lae occurs in the most anterior 1-cm region in the youngest and their synapse formation with mitral and other granule age groups and stabilizes in number with 11–12 year olds cells stabilizes (Hinds and Hinds 1976). As regards behavior and adults, it is not surprising that different locations are although newborn humans respond with adult-like facial more reliable for predicting the overall papillae density of expressions to sweet, sour, and bitter tastants, psychophysical the anterior tongue of different age groups. The finding that data indicate humans do not respond to the taste of salt Area 5 is the best area to use for predicting total papillae (sodium chloride) similarly to adults until about 2 years of age density in 11–12 year olds and adults also indicates that sta- (Beauchamp et al. 1986). Furthermore, at 8–9 years of age, bilization of the distribution of papillae occurs by the age male humans appear to have a poorer sensitivity for sweet, of 11–12 years. Taken together with the findings that 1) the salt, and sour stimuli (James et al. 1997) and a poorer ability size and shape of papillae stabilizes at the age of 11–12 years to perceive the constituents of taste mixtures (Oram et al. and 2) adults and 8–9 year olds have a similar number of 2001). Thus, even though 8–9 year olds have more papillae taste pores per papilla (Segovia et al. 2002), the data provide and more taste buds than adults, these latter data suggest that strong evidence that the peripheral gustatory system is fully neural maturation is not complete at this age. functional by 11–12 years of age. Loss of FP during childhood may also affect somatosensory Although the data from the 8 regional subpopulations of perception on the tongue. FP receive both taste (chorda tym- papillae indicated which subpopulations should be used for pani) and somatosensory (lingual nerve) afferents. Although predicting overall papillae densities on the anterior tongue, the gustatory afferents distribute into taste buds to synapse examination of larger areas, namely, Areas X and Y used by with taste cells, somatosensory afferents distribute around the other workers, was also found to be predictors of papillae taste buds but terminate predominantly in the apex of papillae. density. However, these latter 2 areas produced less reliable It has been reported that humans with fewer papillae tend to

219 Paper 2: Correa et al., 2013 Appendix E 526 M. Correa et al. have poorer spatial acuity (touch) than those who have signifi- Beauchamp GK, Cowart BJ, Moran M. 1986. Developmental changes in cantly more papillae (Essick et al. 2003). For example, humans salt acceptability in human infants. Dev Psychobiol. 19(1):17–25. classified as supertasters because of their ability to taste the Conger AD. 1973. Loss and recovery of taste acuity in patients irradiated to bitter substance PROP at very low concentrations have signifi- the oral cavity. Radiat Res. 53(2):338–347. cantly higher numbers of papillae and superior spatial acuity Delwiche JF, Buletic Z, Breslin PA. 2001. Relationship of papillae number than those classified as poor PROP tasters. Because 8–9 year to bitter intensity of quinine and PROP within and between individuals. olds have a similar number of taste buds per papilla as adults Physiol Behav. 74(3):329–337. but a greater number of papillae (Segovia et al. 2002), it would Essick GK, Chopra A, Guest S, McGlone F. 2003. Lingual tactile acuity, be expected that the loss of papilla they experience would be taste perception, and the density and diameter of fungiform papillae in accompanied by a decrease in their spatial acuity. Such a pro- female subjects. Physiol Behav. 80(2-3):289–302. posal suggests that children up to mid-childhood would be Fast K, Duffy VB, Bartoshuk LM. 2002. New psychophysical insight in eval- expected to have a superior spatial acuity ability to adults. In uating genetic variation in taste. In: Rouby C, Schaal BM, Dubois D, addition, because the somatosensory afferents respond to Gervais R, Holley A, editors. Olfaction, taste and cognition. New York: chemical irritants, it is possible that children would be more Cambridge University Press. p. 391–407. Downloaded from sensitive than adults to “hot” tasting food ingredients such as Hinds JW, Hinds PL. 1976. Synapse formation in the mouse olfactory bulb. chilli, mustard, and curries or “cooling” sensations as occurs I. Quantitative studies. J Comp Neurol. 169(1):15–40. with mints. As yet, these 2 proposals have not been tested. James CE, Laing DG, Oram N. 1997. A comparison of the ability of 8-9-year-old children and adults to detect taste stimuli. Physiol Behav. 62(1):193–197.

Conclusion http://chemse.oxfordjournals.org/ Miller IJ Jr. 1986. Variation in human fungiform taste bud densities among In conclusion, this investigation has shown that children up to regions and subjects. Anat Rec. 216(4):474–482. mid-childhood have a greater number of FP than adults and Miller IJ Jr. 1988. Human taste bud density across adult age groups. J that their papillae are smaller and more uniform in shape. The Gerontol. 43(1):B26–B30. difference in papillae numbers between adults and 7–8-year-old Miller IJ. 1995. Anatomy of the peripheral taste system. In: Doty RL, editor. children appears to be confined to the most anterior region of Handbook of olfaction and gustation. New York: Marcel Dekker. p. the tongue and it is at a location within this region (Area 1) that 521–547. a subpopulation was found that was the best predictor of the Miller IJ, Reedy FE. 1990a. Quantification of fungiform papillae and taste overall number of papillae on the anterior tongue of children pores in living human subjects. Chem Senses. 15:281–294. up to 10 years of age. The finding that a different subpopula- at University of Sydney Library on October 29, 2015 Miller IJ Jr, Reedy FE Jr. 1990b. Variations in human taste bud density and tion of papillae (Area 5) was a better predictor for 11–12 year taste intensity perception. Physiol Behav. 47(6):1213–1219. olds and adults is yet another feature of the data that indicates the organization of papillae and related neural connections in Mistretta CM, Gurkan S, Bradley RM. 1988. Morphology of chorda tympani fiber receptive fields and proposed neural rearrangements during devel- children continues until 11–12 years of age. Importantly, the opment. J Neurosci. 8(1):73–78. subpopulations of papillae found for predicting overall papillae density are smaller than those reported providing a more rapid Moses SW, Rotem Y, Jagoda N, Talmor N, Eichhorn F, Levin S. 1967. A clini- cal, genetic and biochemical study of familial dysautonomia in Israel. Isr procedure for measuring overall papillae numbers. Finally, the J Med Sci. 3(3):358–371. cessation of papillae loss by 9–10 years of age coincides with the cessation of growth of the anterior tongue (Temple et al. Nagai T, Mistretta CM, Bradley RM. 1988. Developmental decrease in size of peripheral receptive fields of single chorda tympani nerve 2002), suggesting that the gross anatomical development of the fibers and relation to increasing NaCl taste sensitivity. J Neurosci. anterior tongue in humans is achieved by mid-childhood. 8(1):64–72. Oram N, Laing DG, Freeman MH, Hutchinson I. 2001. Analysis of taste Funding mixtures by adults and children. Dev Psychobiol. 38(1):67–77. This work was supported by Postgraduate Scholarship from Schiffman SS, Graham BG, Suggs MS, Sattely-Miller EA. 1998. Effect of the Centre for Advanced Food Research (to M.C.). psychotropic drugs on taste responses in young and elderly persons. Ann N Y Acad Sci. 855:732–737. Segovia C, Hutchinson I, Laing DG, Jinks AL. 2002. A quantitative study of Acknowledgements fungiform papillae and taste pore density in adults and children. Brain The authors wish to thank the university students and the children Res Dev Brain Res. 138(2):135–146. and staff of participating schools for their assistance and cooperation. Shahbake M, Hutchinson I, Laing DG, Jinks AL. 2005. Rapid quantitative assessment of fungiform papillae density in the human tongue. Brain References Res. 1052(2):196–201. Arey LB, Tremaine MJ, Monzingo FL. 1935. The numerical and topographi- Smith DV. 1971. Taste intensity as a function of area and concentration: dif- cal relations of taste buds to human circumvallate papillae throughout ferentiation between compounds. J Exp Psychol. 87(2):163–171. the life span. Anat Rec. 64:9–25. Stein N, Laing DG, Hutchinson I. 1994. Topographical differences in sweet- Arvidson K. 1979. Location and variation in number of taste buds in human ness sensitivity in the peripheral gustatory system of adults and children. fungiform papillae. Scand J Dent Res. 87(6):435–442. Brain Res Dev Brain Res. 82(1-2):286–292.

220 Paper 2: Correa et al., 2013 Appendix E Changes in Fungiform Papillae Density 527

Temple EC. 1999. Aspects of the development of the sense of taste in Yackinous CA, Guinard JX. 2002. Relation between PROP (6-n-propylthiou- humans [dissertation]. University of Western Sydney, New South Wales, racil) taster status, taste anatomy and dietary intake measures for young Sydney. men and women. Appetite. 38(3):201–209. Temple EC, Hutchinson I, Laing DG, Jinks AL. 2002. Taste development: Zuniga JR, Chen M, Phillips CL. 1997. Chemosensory and somatosensory regen- differential growth rates of tongue regions in humans. Brain Res Dev eration after lingual nerve repair in humans. J Oral Maxillofac Surg. 55:2–14. Brain Res. 135(1-2):65–70. Zuniga JR, David SH, Englehardt RA, Miller IJ, Schiffman SS, Phillips C. Tepper BJ, Nurse RJ. 1997. Fat perception is related to PROP taster status. 1993. Taste performance on the anterior human tongue varies with Physiol Behav. 61(6):949–954. fungiform taste bud density. Chem Senses. 18:449–460. Downloaded from http://chemse.oxfordjournals.org/ at University of Sydney Library on October 29, 2015

221 Paper 3: Correa et al., 2015 Appendix E Pediatr Nephrol (2015) 30:2003–2010 DOI 10.1007/s00467-015-3131-5

ORIGINAL ARTICLE

Reduced taste function and taste papillae density in children with chronic kidney disease

Maryam Correa1 & David G. Laing1,2 & Ian Hutchinson1 & Anthony L. Jinks3 & Jessica E. Armstrong2 & Gad Kainer2,4

Received: 16 January 2015 /Revised: 8 May 2015 /Accepted: 8 May 2015 /Published online: 5 June 2015 # IPNA 2015

Abstract 7.17, p=0.007) and poorer taste sensitivity than the CC group Background Taste loss may contribute to the loss of appetite (p=0.0272), and the density correlated significantly with in children with chronic kidney disease (CKD) and other se- eGFR (r=0.56, p<0.01). rious medical conditions that result in malnutrition. Tradition- Conclusions Loss of taste in children with CKD is due al methods for measurement of taste loss commonly use aque- to the reduced number of papillae and their taste- ous tastant solutions that can induce nausea, vomiting, or even sensing receptor cells. pain in the mouth. An alternative is to measure fungiform papillae density on the anterior tongue since this correlates Keywords Children . Kidney disease . Taste . Histology . with taste sensitivity. Here we aimed to develop a non- Fungiform papillae invasive method for assessing papillae density on the anterior tongue and to use the method to determine if CKD patients [estimated glomerular filtrate (eGFR<60 ml/min/1.73 m2)] Introduction have a lower density than clinical controls (CC)(eGFR> 89 ml/min/1.73 m2). Recently, in our continuing studies of the effects of diseases Methods Thirty-five healthy adults participated in the devel- on taste loss in children [1–3], we noted that some patients opment of a method, which was assessed by 24 children, 12 of refused to undertake or complete common psychophysical whom were CKD patients and 12 were clinical controls. taste tests that use aqueous taste solutions due to pain on the Results Similar papillae densities were found using invasive tongue, or induction of nausea or vomiting. Affected patients and non-invasive methods (F(1,34)=0.647, p=0.427). The include those suffering from xerostomia (dry mouth syn- 2 CKD group had a significantly lower papillae density (X = drome) as a result of radiation therapy for head and neck cancer, patients who had lesions on the tongue [4–6], and patients with strong food aversions due to disease and * David G. Laing [email protected] CKD patients. A potentially better approach to measuring taste loss in the latter patients is to measure fungiform papillae density on the 1 Children’s Food Research and Education Unit, Centre for Advanced anterior tongue. This measurement is widely accepted as Food Research, College of Science, Technology and Environment, – University of Western Sydney, Locked Bag 1797, Penrith reflecting taste sensitivity of an individual [7 9]. For example, South, NSW 1797, Australia there is a strong correlation between objective anatomical da- 2 School of Women’s and Children’s Health, Medicine, University of ta, namely, papillae density, with data from subjective psycho- NSW, Level 3 Sydney Children’s Hospital, High Street, physical studies using the tastant propylthiouracil. These stud- Randwick, NSW 2031, Australia ies showed that it was possible to classify individuals as 3 School of Psychology and Social Sciences, University of Western supertasters, tasters and non-tasters on the basis of papillae Sydney, Locked Bag 1797, Penrith South 1797, NSW, Australia density [10–13]. Recently, a rapid method requiring only a 4 Department of Nephrology, Sydney Children’s Hospital, High Street, few minutes for measuring papillae density in the anterior Randwick, NSW 2031, Australia tongue was described using a digital camera [14, 15].

222 Paper 3: Correa et al., 2015 Appendix E 2004 Pediatr Nephrol (2015) 30:2003–2010

Importantly, a small region of the tongue was identified where the papillae density is highly correlated with that for the whole anterior tongue. Only a few images of the small region are required to be recorded to establish the papillae density and relative taste sensitivity of an individual. Current techniques to measure papillae density, however, require drying of the tongue and staining with a food dye [14, 16] or 0.5 % methylene blue [10, 11, 17] to facilitate counting of fungiform papillae prior to capturing images of the tongue. Since the actions of drying and staining the tongue can trigger negative behavioral reactions from patients sensitive to touching the tongue, we sought a less invasive method for counting papillae. Accordingly, one aim of the present study was to develop a rapid non-invasive procedure for measuring papillae density on the anterior tongue that avoids drying and using a dye. A second aim was to demonstrate the practicality of the new procedure in a clinical application, namely, to assess whether patients with CKD have a reduced number of fungiform papillae. We have previously re- ported that children with CKD have reduced taste sen- sitivity [1], but it has not been determined whether taste papillae density is also reduced.

Materials and methods

Development of a non-invasive method for measuring papillae density

Thirty-five adult volunteers aged between 20 and 24 years (26 females and nine males; mean age, 21.63±0.27 SE years consented to participate in the study. A digital camera (Canon AF SLR (EOS-1Ds) equipped Fig. 1 Diagram of the procedures in the three stages of tongue with a macro lens (EF 100 mm f/2.8 macro USM and an preparation for photography extension tube (EF25) was used to photograph the tongue. The extension tube enabled close-up images at high magnifi- 0.9 μS), sat on a chair and supported their head by cation to be obtained by increasing the distance between the placing their arm on a table such that their chin pro- front of the lens and the operator. The configuration also in- truded forward. The subject then protruded their tongue cluded a macro ring light (MR-14EX) with two circular flash and held it steady with their lips. At the commencement tubes which consisted of an E-TTL auto-flash. The latter light- of stage 1, a flexible ruler and a fine paintbrush con- ing arrangement was necessary because it enhanced the im- taining red food dye (Roberts, RED 4R124 edible food ages taken at a close distance. color powder) were used to mark the most anterior bor- The experimental procedure was conducted in three stages der of the area to be photographed [15](Fig.1a). A (Fig. 1 and below). At each stage, the tongue of each partici- 10×3-mm-wide piece of filter paper was then placed pant was photographed. All three stages were conducted on on the right side of the anterior tongue to provide a the same day to ensure the same location on the anterior 1-cm scale to calculate the magnification of each image. tongue of a participant was being photographed. The area The red mark remained in position for stages 2 and 3. photographed has been shown to have a papillae density that At each stage, 2–3 images of the anterior tongue, which highly correlates with the total number of fungiform papillae included the marked area, were captured with the digital in the anterior tongue of adults [15]. At each of the camera. During the 1–2-minute rest period between three stages, participants first rinsed their mouth with stages, the 1-cm scale was removed from the tongue deionized water (Milli-Ro-6 Plus System, conductivity and replaced before photography.

223 Paper 3: Correa et al., 2015 Appendix E Pediatr Nephrol (2015) 30:2003–2010 2005

Stage 1: tongue wet and no stain

The tongue was not dried and was not stained prior to pho- tography (Fig. 1a).

Stage 2: tongue dried and no stain used Fig. 2 Image preparation before the fungiform papillae were counted in Subjects protruded their tongue and it was dried with filter stages 1 and 2 (a) and stage 3 (b)showsa28.27-mm2 circle superimposed paper (Whatman’s No. 1). The 1-cm scale was replaced on on the tongue and a 1-cm paper scale the right side of the anterior tongue and 2–3imagesofthe anterior tongue captured (Fig. 1b). Statistical analysis

The total number of fungiform papillae in the circle (papillae Stage 3: tongue dried, stained, and dried 2 density/cm ) was counted for each participant for each of the three stages. The tongue was dried as in stage 2 and a 6-mm-diameter filter ABland–Altman plot was then used to establish the 95 % paper (28.27 mm2) was then dipped in 0.5 % methylene blue limits of agreement between the papillae densities for each (C152105). Using steel forceps, the filter paper was applied pair of stages 1 vs. 2, 1 vs. 3, 2 vs. 3 and three regression above the red mark close to the midline and removed after 3 s analyses determined whether the limits of agreement were (Fig. 1c). The tongue was again dried with filter paper to different between the three stages. prevent the dye from spreading and 2–3tongueimageswere obtained. Comparison of papillae density and taste function in CKD Image analysis and CC patients

All 24 participants (six females and 18 males aged 5–17 years, All images from individuals were first assigned to three groups and each numbered with a 4-digit number. This proce- median=12 years) were patients of the Department of Ne- phrology Outpatients Clinic, Sydney Children’sHospital.A dure ensured that the identities of the participants were un- parent/guardian was present during every experimental proce- known to the experimenter until all the counts were completed and the data were combined for statistical analyses. dure and provided written consent. The children were also asked if they wished to participate. The participants were divided into two groups based on Stage 1 and 2 analyses their estimated glomerular filtration rate (eGFR) value, which was estimated using the Schwartz formula [18]. One group 2 The images were downloaded to a computer (Toshiba Satellite (CKD patients) had a eGFR<60 ml/min/1.73 m and the other 2 PRO6100) and examined using an Adobe Photoshop 7.0 pro- (CC, clinical controls) a eGFR>89 ml/min/1.73 m . Patients gram. The midline of the tongue was drawn on the image with CKD have been reported to have taste loss [1] and less using the Bgrid^ line from the Bview^ menu. The 1-cm scale taste buds per papillae [19], but it has not been determined on each image was used to calculate and superimpose a 28.27- whether papillae density is also reduced. mm2 circle on the image where it was marked with the red The patient and control groups were age and gender food dye. The image was then magnified to enhance identifi- matched with two exceptions where a female and male of cation and counting of the total number of fungiform papillae the same age were matched (Table 1). Clinical controls instead in the circle (Fig. 2a) of healthy controls were included in the study because of their availability and because we had found there were no signifi- cant differences between taste function in healthy and clinical Stage 3 analysis controls [1]. Furthermore, no previous studies of children or adults with various types of renal disease or early stages of After the images of the tongue were downloaded into the CKD have been found to exhibit improved taste function. computer, the midline of the image was marked as in the latter Accordingly, it was assumed that inclusion of patients with analyses. After superimposing the circle on the stained area, renal disease in the control group would not reduce the chance the images were magnified and the fungiform papillae inside of finding a difference in taste function when compared to the circle were counted (Fig. 2b). CKD patients. The researcher who photographed the tongues

224 Paper 3: Correa et al., 2015 Appendix E 2006 Pediatr Nephrol (2015) 30:2003–2010

Table 1 Participant characteristics and diagnoses

Patient Age M/F# GFR Diagnosis Control Age M/F# GFR Diagnosis

1** 14 M 56.7 PUV* 1 15 F 161 MCGN* 2 14 M 58 Dysplastic 2 15 M 141 HSP* 3 16 M 20 Reflux nephropathy 3 16 M 94 Hypertension 4 12 M 34 PUV 4 11 M 94 Nephrotic 5 6 M 32 Dysplastic 5 7 M 127 Hypertension 6 16 M 25 Reflux nephropathy 6 18 M 104 RPGN* 7 6 M 47 Dysplastic 7 8 M 97 Post BMT* nephropathy 8 17 M 11.6 Dysplastic 8 17 M 101 FSGS* 9 11 F 15 Dysplastic 9 11 F 112 UTI* 10 15 M 9.6 Dysplastic 10 15 M 174 FSGS 11 9 F 14 Dysplastic 11 8 F 125 Reflux nephropathy 12** 6 M 23 Dysplastic 12 5 F 113 Reflux nephropathy

# M male, F female * BMT bone marrow transplant; FSGS focal segmental glomerulosclerosis; HSP Henoch-Schönlein purpura; MCGN mesangiocapillary glomerulone- phritis; PUV posterior urethral valves; RPGN rapidly progressive glomerulonephritis; UTI urinary tract infection ** Gender difference was unaware of the children’s eGFR and underlying medical difference between the taste identification scores of the condition at the time of the study. CKD and CC groups.

Procedures Results The equipment, experimental procedure, location on the tongue, and image analysis were as described in stage 1. No Development of a non-invasive method for measuring food dye was used for staining papillae, nor were the papillae papillae density dried. For each subject, fungiform papillae were counted in the superimposed 28.27-mm2 circle and the magnification Description of fungiform papillae in each of the three stages recorded. The appearance of fungiform papillae was different in the three treatment stages. In stage 1, the papillae were elevated Taste function assessment bright pink and red structures, which were larger than the surrounding filiform papillae and easily distinguished from Taste function was assessed using a three-choice taste identi- the background surface of the tongue (Fig. 3a). fication procedure described earlier [1], which involved the In stage 2, the fungiform papillae were pale pink in color identification of five concentrations of aqueous solutions of and some of the smaller papillae were flattened in comparison sweet, sour, salty, and bitter tastants, which ranged from very to their appearance in stage 1 (Fig. 3b). In some instances it strong to very weak. Each child was assessed on the same day was difficult to identify the smaller fungiform papillae be- that their papillae were counted. cause they blended with the surrounding filiform papillae. In stage 3, the fungiform papillae were elevated similar to Statistical analyses stage 1, but not all the papillae were stained with methylene

Two Kruskal–Wallis tests followed by post hoc Mann–Whit- ney tests were conducted to investigate whether there was a significant difference in fungiform papillae densities and eGFR values between the CKD and CC groups. A Pearson correlation (two-tailed) analysis was conducted to investigate whether was a relationship between the eGFR and the fungi- ’ form papillae density of all the patients and a Fisher s Fig. 3 Fungiform papillae of a participant in stage 1 (a), stage 2 (b), and exact test determined whether there was a significant stage 3 (c)

225 Paper 3: Correa et al., 2015 Appendix E Pediatr Nephrol (2015) 30:2003–2010 2007 blue (Fig. 3c). Some fungiform papillae stained darker blue there was a significant correlation between the fungiform pa- whereas, and others were only slightly stained. However, in all pillae density and eGFR (r=0.56, p=0.004), with the number images the surrounding tissue was always darker in color than of papillae decreasing significantly with lower eGFR (Fig. 4). in stage 2 because the surrounding tissue and the filiform Fisher’s exact test indicated that the CC group recorded a papillae were stained darker. significantly higher taste identification score than the CKD group (p=0.027). In addition, a Mann–Whitney test showed Papillae density in the three stages that there was a significant difference in papilla density be- tween patients whose taste sensation was diminished com- The median and range of fungiform papillae density for im- pared to those with unaffected taste sensation (U=28.5, medi- ages from stages 1, 2, and 3, respectively, were 1.38/cm2 an difference in papillae density 7.07/cm2), (95 % CI 0–17.7, (0.38–2.38), 1.38/cm2 (0.38–2.13) and 1.50/cm2 (0.38– p<0.05) 2.63). Bland–Altman plots indicated that the 95 % limits of agreement (LoA)(−3.06, 2.94) contained 94 % (33/35) of the Taste assessment difference scores (papillae densities) for stages 1 and 2, 97 % (34/35), (LoA: − 4.94, 2.54) and 97 % (34/35), (LoA: − 4.16, Significantly more CKD children (7/12) exhibited a decrease 1.87) for stages 1 and 3 and stages 2 and 3, respectively. In in their taste identification ability than the CC group (1/12) addition, regression analyses indicated there were no signifi- (p=0.0272) (Table 2). cant differences between stages 1 and 2 (p=0.766)stages1 In summary, the three most important findings were that and 3 (p=0.108) and stages 2 and 3 (p=0.100), indicating that fungiform papillae density and taste identification ability of there was a significant level of agreement between the papillae CKD children were significantly lower than in the controls densities measured using the three methods. and the papillae counts decreased with decreasing eGFR Thus, the analyses showed that there was no significant values. difference in our ability to identify and count fungiform papil- lae using each of the above three methods. Importantly, it was possible to identify the fungiform papillae from the surround- Discussion ing filiform papillae in stages 1 and 2, even though in stage 2 some of the smaller fungiform papillae were more difficult to The study produced four major outcomes; (1) it demonstrated identify because they had been flattened during the drying that CKD patients had a significantly lower papillae density process (Fig. 3b). In stage 3, it was simpler to identify fungi- than the CC group; (2) taste loss was more prevalent in CKD form papillae from the surrounding filiform papillae than from patients than the controls; (3) a strong correlation between stage 2 because the fungiform papillae were stained. Overall, eGFR and papillae density was found; and (4) a new method the results indicated that it is possible to quantify fungiform was described which allowed quantification of fungiform pa- papillae density in a region of the anterior tongue without pillae density on the anterior human tongue without using using a staining or drying procedure. Accordingly, the method invasive staining and drying procedures. used in stage 1 was used to measure papillae density of the two The finding that the new non-invasive photographic tech- patient groups. nique differentiated between patients with CKD and controls on the basis of papillae numbers represents the first time this A comparison of papillae density and taste function has been reported. It complements the report that indicated in CKD and CC patients 78 % of the fungiform papillae in CKD patients did not con- tain taste buds in comparison to only 12 % of the papillae of The median papillae density for the CC group was 4.00 papil- controls [19]. The present and past data therefore indicate that lae/cm2 with a range of 0.88 to 6.64 papillae/cm2, while the CKD patients have lower numbers of papillae and taste buds, CKD group median was 2.4 papillae/cm2 with a range of 1.77 which would lower the number of taste receptor cells within to 3.98 papillae/cm2. these entities and decrease their ability to perceive tastes. This The medians and ranges of eGFR for the CC and CKD prediction was found to be true in the present study where groups, respectively, were 112.5, 94 to 174 and 24, 9.6 to significantly greater numbers of CKD patients than controls 58 ml/min/1.73 m2. Kruskal–Wallis tests showed that both had a poorer ability to perceive common tastes. A similar fungiform papillae density (X2=7.17, p=0.007) and eGFR finding was made in a larger psychophysical study that includ- (X2=17.28, p<0.001) were significantly less in the CKD ed some of the children in the present study [1]. In addition, a group than in the CC group. The post hoc Mann–Whitney number of studies have reported that people suffering from tests also confirmed that there was a significant difference renal failure have impaired taste acuity [20–22]. In several between the CC and CKD groups for papillae density (U= studies, patients with renal disease have exhibited raised 26.0, p=0.0068)andeGFR(U=0.00,p<0.001). Furthermore, thresholds for the primary tastes (sweet, sour, salt, bitter)

226 Paper 3: Correa et al., 2015 Appendix E 2008 Pediatr Nephrol (2015) 30:2003–2010

200

) 150 2 m37.1/nim/Lm(erocSRFGe

100

50

0 0102030405060 Fungiform Papillae Density (cm2) Fig. 4 Relationship between estimated glomerular filtration rate (eGFR) individuals, respectively. Fungiform papillae density scores represent and fungiform papillae density. Filled and unfilled diamonds represent the the mean number of papillae per 28.27 mm2 circle results of clinical control (CC) and chronic kidney disease (CKD)

[23, 24]. Nevertheless, in the present study, although more was supported by a larger earlier psychophysical study [1] CKD patients had taste loss than those in the CC group, the where a strong association was also reported between taste variation in papillae density between participants was too loss as measured by the ability to identify tastes and eGFR, large and the sample size too small to allow us to nominate a with decreasing values of eGFR correlating with a poorer specific papillae density below which taste loss would occur. ability to identify sweet, bitter, and sour tastes. However, the Importantly, however, the methods and results described here small numbers of patients and differences in underlying etiol- for measuring papillae density provide a simple way for fol- ogy of renal diseases between the CKD group and those with lowing the loss and recovery of taste function in individual milder renal impairment (CC group) precludes a definite con- patients with CKD, various cancers, and other diseases that clusion that CKD causes reduced taste and papilla density. are known to affect taste function. Nevertheless, a reduced ability to discern taste was associated Another new finding in the study was the strong associa- significantly with reduced papilla density. It should be noted tion between eGFR and papillae density (Fig. 4). This finding that the etiology of renal diseases in all but one of the patients

Table 2 Papillae density and 2 2 taste loss Patient Papillae/ Papillae/cm Taste loss Control Papillae/ Papillae/cm Taste loss circle circle

1 4 14.15 Yes 1 12 42.445 No 2 9 31.84 Yes 2 15 53.06 No 3 6 21.22 No 3 2 7.07 No 4 6 21.22 No 4 5 17.69 Yes 5 4 14.15 Yes 5 10 35.37 No 6 7 24.76 No 6 11 38.91 Not tested 7 5 17.69 No 7 14 49.52 No 8 5 17.69 Yes 8 8 28.30 No 9 5 17.69 No 9 7 24.76 No 10 5 17.69 Yes 10 8 28.30 No 11 6 21.22 Yes 11 7 24.76 No 12 7 24.76 Yes 12 12 42.45 No

227 Paper 3: Correa et al., 2015 Appendix E Pediatr Nephrol (2015) 30:2003–2010 2009 in the CKD group was congenital dysplasia compared with anterior tongue that are of sufficient clarity to allow quantifi- only two patients with congenital dysplasia in the control cation of papillae density in a small region, which has been group. The length of time that patients were exposed to dete- shown to reflect the papillae density of the whole anterior riorating renal function or the etiology of CKD may also have tongue [15]. As mentioned above, this is important because had a bearing on the papilla density and taste dysfunction. papillae density provides a qualitative measure of taste sensi- Longitudinal studies to assess papilla density over time and tivity changes in humans [7–9]. The newly described method, progression of CKD may clarify the relationship between which is relatively simple to perform in a clinical setting, has CKD etiology, disease progression, and taste dysfunction. Al- the potential to indicate whether loss or recovery of taste is though the mechanisms underlying taste loss in CKD patients occurring during treatment, recovery, or deterioration from are not known, it has been suggested that changes in the levels chronic illness in individual patients. Another practical appli- of Na and Ca that characterize CKD could impede transduc- cation of importance is that data obtained required only a few tion processes at receptor cells in papillae and interfere with images from each patient and a few minutes of their time, taste function [25]. It is suggested here that interference with minimizing any inconvenience to a patient. Indeed, the meth- the taste reception process could lead to decreased receptor od has been used at the bedside of hospitalized patients, out- turnover and result in papillae loss through lack of stimulation. patients in clinics, schools, and in research laboratories. Indeed the fact that lower numbers of taste buds were found in Concerning the use of the non-staining, non-drying method fungiform papillae of CKD patients [19] suggests that changes in cases where a subject cannot tolerate the 1-cm paper scale at the receptor level may have occurred. However, despite the being placed on their tongue and/or the very small amount of relationships between taste function and papillae loss with red marker dye, an alternative is to use the digital grid method reduced eGFR, there is no evidence at present that low eGFR [15]. The latter method can be used to identify the location and is the direct cause of papillae or taste loss. size of the 28.27-mm2 circle on the anterior tongue. Using this The finding that there were no significant differences be- method, the mean error for calculating the diameter of the tween the numbers of papillae using the three quantitation circle and location of the circle is<6 % (based on 35 partici- procedures indicates that it is possible to measure papillae pants) (Correa, unpublished data) compared to using the 1-cm density without any disturbance to the surface of the tongue scale; an acceptable compromise. by dyes or drying traditionally used in such studies. Given the In conclusion, the present study showed that it is unneces- array of clinical disorders and treatments that can damage the sary to use histological stains or drying procedures to deter- tongue’s sensory apparatus, this finding has applications be- mine fungiform papillae density on the anterior tongue of yond its use in measuring reduced taste function in CKD pa- humans. In conjunction with a digital camera, the method tients. Furthermore, none of the children in the present study used here provides a rapid non-invasive procedure for funda- found the technique unpleasant and none refused to allow the mental studies of taste anatomy and function on the anterior measurements to be undertaken. The reasons for promoting tongue and for monitoring taste loss and recovery in patients this new (no dye or drying) non-invasive method for assessing with diseases and medications that can affect taste. As regards papillae are twofold. First, it was easier to distinguish fungi- CKD patients, the evidence found here for reduced papillae form papillae when the tongue surface was not dried, as the density, reduced taste function and an association of papillae papillae were more elevated. When the tongue was dried, density with reduced eGFR supports the earlier finding of an some of the fungiform papillae were flattened, making their association between taste and renal dysfunction [1]. The identification more difficult. Second, historically, the develop- changes in taste function support the long-held view that ment of methods for studying fungiform papillae have changes in the sense of taste of CKD patients may be one of evolved from histopathological techniques that used light the underlying factors that affects the eating behavior of these and electron microscopy of animal and cadaver tongues to patients resulting in their loss of appetite, altered food study papillae and taste buds where structural features of pa- preferences, aversion to particular foods, and poor pillae were often the focus. For example, all recent studies of growth. Accordingly, identification of taste loss, and in papillae density have used either methylene blue or blue food particular which tastes are most affected during CKD, dyes to view papillae, arguing that this allowed the papillae to could result in improved food intake, management of be better distinguished from the surrounding filiform papillae. patients during treatment, and better outcomes as Furthermore, drying the tongue with paper became standard regards responding to treatments. practice to avoid the dye spreading beyond areas of interest and to enhance staining. Surprisingly, no one appears to have Acknowledgments The authors wish to thank the adults and children compared conditions where the drying and staining were com- who participated in the study for their assistance and cooperation and the ’ pared with the natural condition of the tongue. Thirdly, the Sydney Children s Hospital Foundation for a grant supporting the re- search (DGL and GK). MC was supported by a Postgraduate Scholarship results demonstrated that a moderately high resolution digital from the Centre for Advanced Food Research and JEA by an Australian camera can provide images of papillae in the unstained Postgraduate Award.

228 Paper 3: Correa et al., 2015 Appendix E 2010 Pediatr Nephrol (2015) 30:2003–2010

Compliance with ethical standards All procedures performed in stud- 12. Bartoshuk LM, Duffy VB, Miller IJ (1994) PTC/PROP tasting: ies involving human participants were in accordance with the ethical anatomy, psychophysics and sex effects. Physiol Behav 56:1165– standards of the institutional and/or national research committee and with 1171 the 1964 Declaration of Helsinki and its later amendments or comparable 13. Prutkin J, Fisher EM, Etter L, Fast K, Gardner E, Lucchina LA, ethical standards. Snyder DJ, Tie K, Weiffenbach J, Bartoshuk LM (2000) Genetic variation and inferences about perceived taste intensity in mice and Conflict of interest The authors declare they have no conflicts of men. Physiol Behav 69:161–173 interest. 14. Shahbake M, Hutchinson I, Laing DG, Jinks AL (2005) Rapid quantitative assessment of fungiform papillae density in the human tongue. Brain Res 1052:196–201 15. Correa M, Hutchinson I, Laing DG, Jinks AL (2013) Changes in References fungiform papillae density during development in humans. Chem Sens 38:519–527 16. Fast K, Duffy VB, Bartoshuk LM (2002) New psychophysical 1. Armstrong JE, Laing DG, Wilkes FJ, Kainer G (2010) Smell and insight in evaluating genetic variation in taste. In: Rouby C, taste function in children with chronic kidney disease. Pediatr Schaal B, Dubois D, Gervais R, Holley A (eds) Olfaction, taste Nephrol 25:1497–1504 and cognition. Cambridge University Press, New York, pp 391–407 2. Laing DG, Wilkes FJ, Underwood N, Tran L (2011) Taste disorders 17. Segovia C, Hutchinson I, Laing DG, Jinks AL (2002) A quantita- in Australian Aboriginal and Non-Aboriginal children. Acta Pediatr tive study of fungiform papillae and taste pore density in adults and 100:1267–1271 children. Dev Brain Res 138:135–146 3. Cohen J, Laing DG, Wilkes FJ (2012) Taste and smell function in 18. Schwartz G, Haycock G, Eldermann CJ, Spitzer A (1976) A simple pediatric blood and bone marrow patients. Support Care Cancer 20: estimate of glomerular filtration rate in children derived from body 3019–3023 length and plasma creatinine. Pediatrics 58:259–263 4. Henkin RI, Talal N, Larson AL, Mattern CFT (1972) Abnormalities 19. Astback J, Fernstrom A, Hylander B, Arvidson K, Johansson O of taste and smell in Sjögren's syndrome. Ann Int Med 76:375–383 (1999) Taste buds and neuronal markers in patients with chronic – 5. Mossman KL (1983) Quantitative radiation dose–response relation- renal failure. Perit Dial Int 19:s315 s323 20. Fernstrom A, Hylander B, Rossner S (1996) Taste acuity in patients ships for normal tissue in man. II. Response of the salivary glands – during radiotherapy. Radiation Res 95:392–398 with chronic renal failure. Clin Nephrol 45:169 174 21. Middleton RA, Allman-Farinelli MA (1999) Taste sensitivity is 6. Sandow PL, Hejrat-Yazdi M, Heft MW (2006) Taste loss and re- altered in patients with chronic renal failure receiving continuous covery following radiation therapy. J Dental Res 85:608–611 ambulatory peritoneal dialysis. J Nutrit 129:122–125 7. Delwiche JF, Buletic Z, Breslin PA (2001) Relationship of papillae 22. Matson A, Wright M, Oliver A, Woodrow G, King N, Dye L, number to bitter intensity of quinine and PROP within and between Blundell J, Brownjohn A, Turney J (2003) Zinc supplementation – individuals. Physiol Behav 74:329 337 at conventional doses does not improve the disturbance of taste 8. Smith DV (1971) Taste intensity as a function of area and concen- perception in hemodialysis patients. J Renal Nutrit 13:224–228 tration: differentiation between compounds. J Exp Psychol 23. Atkin-Thor E, Goddard BW, O'Nion J, Stephen RL, Kolff WJ – 87:163 171 (1978) Hypogeusia and zinc depletion in chronic dialysis patients. 9. Stein N, Laing DG, Hutchinson I (1994) Topographical differences Am J Clin Nutr 31:1948–1951 in sweetness sensitivity in the peripheral gustatory system of adults 24. Burge JC, Park HS, Whitlock CP, Schemmel RA (1979) Taste and children. Dev Brain Res 82:286–292 acuity in patients undergoing long term hemodialysis. Kidney Int 10. Miller IJ, Reedy FE (1990) Quantification of fungiform papillae 15:49–53 and taste pores in living human subjects. Chem Sens 15:281–294 25. Zufall F, Firestein S, Shepherd GM (1994) Cyclic nucleotide-gated 11. Miller IJ, Reedy FE (1990) Variation in human taste bud density ion channels and sensory transduction in olfactory sensory neurons. and taste sensitivity. Physiol Behav 47:1213–1219 Annu Rev Biophys Biomol Struct 23:577–607

229