Non-Nutritive Sweetened Beverages and Cardiometabolic Risk

Non-Nutritive Sweetened Beverages and Cardiometabolic Risk

by

Néma McGlynn

A thesis submitted in conformity with the requirements for the degree Master of Science

Department of Nutritional Sciences University of Toronto

© Copyright by Néma McGlynn 2020

ii

Non-Nutritive Sweetened Beverages and Cardiometabolic Risk

Néma McGlynn

Master of Science

Department of Nutritional Sciences

University of Toronto

2020

Abstract

Whether non-nutritive sweetened beverages (NSBs) improve cardiometabolic risk factors similar

to water in their intended substitution for sugar sweetened beverages (SSBs) is unclear. To assess

the effect of 3 prespecified substitutions (NSBs for SSBs, NSBs for water and water for SSBs)

on body weight and cardiometabolic risk factors, we undertook a systematic review and network

meta-analysis. Fourteen trials (n=1530) predominantly in people with overweight/obesity at risk

for or with diabetes were identified. The substitution of NSBs for SSBs improved body weight,

BMI, body fat, triglycerides and intrahepatocellular lipids supporting the use of NSBs, as a

viable alternative to water, for SSBs. Due to imprecision in the estimate and concerns that NSBs

adversely affect glucose tolerance through compositional gut microbiome changes, we provided

a rationale, design and baseline characteristics report for the STOP Sugars NOW trial to

determine the effect of our 3 prespecified substitutions on gut microbiome and glucose tolerance.

Words: 150

iii

Acknowledgments

I would sincerely like to thank the following individuals whose support and guidance were

pivotal in the completion of this research and attainment of my master’s degree!

First of all, to my sister Natalie, for putting up with losing her spot at the dining room table to

two computer screens, and gently reminding me when my random notes started taking over the

living room. I know it was driving you crazy little cat so thank you for biting your tongue! To

my mother and father for supporting my decision to leave full time employment to pursue

graduate studies. Your Costco deliveries of toilet paper, paper towel and nuts were a lifesaver!

To the Toronto 3D lab crew, including the MSBers, I want to thank each of you for answering all

my questions, even if they were ridiculous. I always looked forward to coming into the lab and

working with you all. Andrea Glenn, Andreea Zarbau, Rodney (little big bro) Au-Yeung,

Meaghan Kavanagh, Jarvis Noronha, Catherine Braunstein, Dr. Laura Chiavaroli, Sonia Blanco

Mejia, Maxine Seider, Sabrina Ayoub-Charette, Stephanie Nishi, Effie Viguiliouk, Danielle Lee,

Annette Cheung, Amna Ahmed, Darshna Patel, Sandhya Sahya-Pudaruth, Melanie Paquet,

Bashyam Balachandran and Stefan Kabisch you all made this journey so much fun!

Dr. David Jenkins, our conversations during my weekends in the lab were a HUGE pick-me-up. I

always loved hearing your thoughts on the many topics we discussed. You were my favourite

part of the day.

A BIG thank you to my committee members Dr. Elena Comelli and Dr. Tom Wolever. Your

advice and direction both in between and during my committee meetings really helped me stay

focused and on track. I would also like to acknowledge Dr. Anthony Hanley, first for guiding me

through your nutritional epidemiology course, it was a rough ride but truly rewarding, and for

offering your time to be my examiner. You have a very positive reputation among the students in

the Nutritional Sciences department and now I know why!

A special acknowledgment to Louisa Kung, our graduate administrator, who went above and

beyond to help me on several different occasions. You are the backbone of the nutritional

sciences department and someone I could always count on.

iv

Dr. Tauseef Ahmad Khan, words cannot express the gratitude I feel towards your endless support

and guidance in every aspect of my research projects. I leaned on you a lot and will be forever

grateful for your patience and understanding during my fumbles through this process. You are a

wonderful educator who really knows how to get the most out of the “copy paste” method.

Finally, a very special thank you to the one and only Dr. John Sievenpiper. Your endless drive

and determination have given so many students an opportunity to learn from the best! Thank you

for pushing me out of my comfort zone. Your unwavering dedication to upholding scientific

rigor in search of the “truth” despite ALL the obstacles that have been thrown your way is

incredibly inspirational. I don’t know how you do it but I’m incredibly proud to say that you

were my supervisor. YOU are the trailblazer.

v

Table of Contents Acknowledgments.......................................................................................................................... iii

Table of Contents ............................................................................................................................ v

List of Tables ................................................................................................................................. ix

List of Figures ................................................................................................................................ xi

List of Abbreviations .................................................................................................................... xv

Chapter 1: Introduction ................................................................................................................... 1

Chapter 2: Literature Review .......................................................................................................... 3

2.1 Obesity and Diabetes ................................................................................................................ 3

2.1.2 Overview of Obesity and Diabetes..................................................................................... 3

2.1.3 Prevalence and Economic Burden...................................................................................... 3

2.1.4 Risk Factors and Complications ......................................................................................... 3

2.1.5 Measures of Overweight and Obesity ................................................................................ 5

2.1.6 Diagnosis and Targets of Control for Type 2 Diabetes ...................................................... 6

2.1.7 Prevention and Management .............................................................................................. 7

2.2 Sugar-Sweetened Beverages ..................................................................................................... 7

2.2.1 Worldwide Intakes of Sugar-Sweetened Beverages .......................................................... 9

2.2.2 Guidelines for Sugar-Sweetened Beverage Consumption ............................................... 10

2.2.3 Replacements for Sugar-Sweetened Beverages in Guidelines ......................................... 10

2.3 Non-Nutritive Sweeteners ....................................................................................................... 13

2.3.1 Definition of Non-Nutritive Sweeteners .......................................................................... 13

2.3.2 Safety and Approval Process............................................................................................ 13

2.3.3 Sweetening Intensity ........................................................................................................ 16

2.3.4 Non-Nutritive Sweetener Consumption Trends ............................................................... 16

2.3.5 Absorption, Digestion, Metabolism and Excretion of Non-Nutritive Sweeteners ........... 16

2.3.5.1 Aspartame .................................................................................................................. 16

2.3.5.2 Sucralose .................................................................................................................... 17

2.3.5.3 Acesulfame Potassium and Saccharin ....................................................................... 17

2.3.6 Potential Mechanisms of Non-Nutritive Sweeteners on Body Weight and Glucose

Control ....................................................................................................................................... 19

2.3.6.1 Activation of Sweet Taste Receptors ......................................................................... 19

2.3.6.2 Changes in Taste Preferences and Dietary Intake ..................................................... 19

2.3.6.3 Gut Microbiome ........................................................................................................ 20

vi

2.3.7 Systematic Reviews and Meta-Analyses of Prospective Cohort Studies Investigating the

Association between Non-Nutritive Sweeteners and Obesity/ Type 2 Diabetes Mellitus Risk 21

2.3.8 Systematic Reviews and Meta-Analyses of Randomized Controlled Trials Investigating

the Effect of Non-Nutritive Sweeteners on Obesity/ Type 2 Diabetes Mellitus Risk............... 22

2.3.9 Network Meta-Analyses ................................................................................................... 26

Chapter 3: Rationale and Objectives............................................................................................. 27

3.1 Rationale.............................................................................................................................. 27

3.2 Objectives ............................................................................................................................ 27

Chapter 4: Effect of Non-Nutritive Sweetened Beverages as a Replacement Strategy for Sugar-

Sweetened Beverages on Body Weight and Cardiometabolic Risk: A Network Meta-Analysis of

Randomized Controlled Trials ...................................................................................................... 29

4.1 Abstract ................................................................................................................................... 30

4.2 Introduction ............................................................................................................................. 31

4.3 Methods................................................................................................................................... 32

4.3.1 Design............................................................................................................................... 32

4.3.2 Data Sources and Searches ............................................................................................... 32

4.3.3 Study Selection ................................................................................................................. 33

4.3.4 Data Extraction ................................................................................................................. 33

4.3.5 Outcomes .......................................................................................................................... 33

4.3.6 Risk of Bias Assessment .................................................................................................. 34

4.3.7 Data Synthesis .................................................................................................................. 34

4.3.8 Grading of the Evidence ................................................................................................... 35

4.3.9 Patient Involvement.......................................................................................................... 36

4.4 Results ..................................................................................................................................... 37

4.4.1 Search Results .................................................................................................................. 37

4.4.2 Available Data .................................................................................................................. 37

4.4.3 Trial Characteristics ......................................................................................................... 37

4.4.4 Risk of Bias ...................................................................................................................... 38

4.4.5 Effect of Substitution of NSBs for SSBs ......................................................................... 38

4.4.6 Effect of Substitution of Water for SSBs ......................................................................... 38

4.4.7 Effect of Substitution of NSBs for Water ........................................................................ 39

4.4.8 Inconsistency .................................................................................................................... 39

4.4.9 Intransitivity (a domain of indirectness, in the indirect estimates) .................................. 39

vii

4.4.10 Sensitivity Analyses ....................................................................................................... 39

4.4.11 Subgroup Analyses ......................................................................................................... 39

4.4.12 Publication Bias.............................................................................................................. 40

4.4.13 Grading of the Evidence ................................................................................................. 40

4.5 Discussion ........................................................................................................................... 40

4.5.1 Findings in the Context of Existing Studies ..................................................................... 41

4.5.2 Strengths and Limitations................................................................................................. 43

4.5.3 Implications ...................................................................................................................... 46

4.5.4 Conclusions ...................................................................................................................... 46

4.5.5 Funding Statement............................................................................................................ 47

4.5.6 Conflict of Interest ........................................................................................................... 48

4.5.7 Ethics Approval ................................................................................................................ 51

4.5.8 Data Sharing ..................................................................................................................... 51

Chapter 5: Rationale, Design and Baseline Characteristics Assessing the Effect of Substituting

NSBs versus Water for SSBs on Gut Microbiome, Glucose Tolerance, and Cardiometabolic Risk

Factors: Strategies To OPpose SUGARS with Non-nutritive sweeteners Or Water trial (STOP

Sugars NOW) ................................................................................................................................ 98

5.1 Abstract ................................................................................................................................... 99

5.2 Introduction ........................................................................................................................... 100

5.3 Objective and Hypotheses..................................................................................................... 101

5.4 Methods................................................................................................................................. 102

5.4.1 Study Design .................................................................................................................. 102

5.4.2 Blinding .......................................................................................................................... 103

5.4.3 Participants ..................................................................................................................... 103

5.4.4 Sample Size (Power) Calculation ................................................................................... 104

5.4.5 Recruitment, Consent and Screening ............................................................................. 105

5.4.6 Run-In Phase .................................................................................................................. 106

5.4.7 Randomization ............................................................................................................... 106

5.4.8 Intervention .................................................................................................................... 106

5.4.9 Study Visits .................................................................................................................... 107

5.4.10 Washout phase.............................................................................................................. 108

5.4.11 Outcomes ...................................................................................................................... 109

5.4.12 Adherence Assessment ................................................................................................. 109

viii

5.4.13 Analytical Techniques .................................................................................................. 109

5.4.13.1 Anthropometric Analyses ...................................................................................... 109

5.4.13.2 Biochemical Analyses ........................................................................................... 110

5.4.13.3 Biomarkers of Adherence ...................................................................................... 111

5.4.14 Antibiotic Use .............................................................................................................. 111

5.4.15 Compensation ............................................................................................................... 112

5.4.16 Statistical Analyses ...................................................................................................... 112

5.4.16.1 Subgroup Analysis ................................................................................................. 112

5.4.17 Adverse Effects ............................................................................................................ 112

5.5 Results ................................................................................................................................... 113

5.5.1 Participant Flow ............................................................................................................. 113

5.5.2 Baseline Characteristics ................................................................................................. 113

5.6 Discussion ............................................................................................................................. 114

5.7 Conclusion ............................................................................................................................ 115

5.8 Acknowledgements and Funding .......................................................................................... 116

5.9 Statement of Contribution ..................................................................................................... 116

Chapter 6: General Discussion.................................................................................................... 125

6.1 Summary ............................................................................................................................... 125

6.2 Strengths and Limitations ..................................................................................................... 126

6.3 Clinical Implications ............................................................................................................. 128

6.4 Future Directions .................................................................................................................. 129

Chapter 7: Conclusions ............................................................................................................... 130

References ................................................................................................................................... 132

ix

List of Tables

Chapter 2

Table 2.1 Harmonized Criterion for the Diagnosis of Metabolic Syndrome[35] ............................. 4

Table 2.2 BMI Classifications[36] ................................................................................................... 5

Table 2.3 Diagnostic Criterion for Type 2 Diabetes Mellitus [12] .................................................. 7

Table 2.4 Non-Nutritive Sweetener Guideline Recommendations .............................................. 12

Table 2.5 Common Non-Nutritive Sweeteners Used in Foods and Beverages in Canada[82, 91, 92]

....................................................................................................................................................... 15

Table 2.6 Systematic Reviews and Meta-Analyses of Randomized Controlled Trials

Investigating the Effect of Non-Nutritive Sweeteners on Adiposity ............................................ 23

Table 2.7 Systematic Reviews and Meta-Analyses of Randomized Controlled Trials

Investigating the Effect of Non-Nutritive Sweeteners on Diabetes Risk ..................................... 25

Chapter 4

Table 4.1: Trial Characteristics .................................................................................................... 52

Appendix Table 4.1: PRISMA-NMAa Checklist ........................................................................ 58

Appendix Table 4.2: Search Strategy.......................................................................................... 62

Appendix Table 4.3: PICOTSb Framework ................................................................................ 63

Appendix Table 4.4: Side-Splitting Approach for Inconsistency for Body Weight ................... 64

Appendix Table 4.5: Side-Splitting Approach for Inconsistency for BMI ................................. 64

Appendix Table 4.6: Side-Splitting Approach for Inconsistency for Body Fat % ..................... 64

Appendix Table 4.7: Side-Splitting Approach for Inconsistency for Waist Circumference ...... 64

Appendix Table 4.8: Side-Splitting Approach for Inconsistency for HbA1c ............................. 65

Appendix Table 4.9: Side-Splitting Approach for Inconsistency for FPG ................................. 65

Appendix Table 4.10: Side-Splitting Approach for Inconsistency for 2h-PG ............................ 65

Appendix Table 4.11: Side-Splitting Approach for Inconsistency for FPI ................................. 65

Appendix Table 4.12: Side-Splitting Approach for Inconsistency for HOMA-IR ..................... 66

Appendix Table 4.13: Side-Splitting Approach for Inconsistency for LDL-C ........................... 66

Appendix Table 4.14: Side-Splitting Approach for Inconsistency for Non-HDL-C .................. 66

Appendix Table 4.15: Side-Splitting Approach for Inconsistency for TGs................................ 66

Appendix Table 4.16: Side-Splitting Approach for Inconsistency for HDL-C .......................... 67

x

Appendix Table 4.17: Side-Splitting Approach for Inconsistency for TC ................................. 67

Appendix Table 4.18: Side-Splitting Approach for Inconsistency for SBP ............................... 67

Appendix Table 4.19: Side-Splitting Approach for Inconsistency for DBP ............................... 67

Appendix Table 4.20: Side-Splitting Approach for Inconsistency for IHCL ............................. 68

Appendix Table 4.21: Side-Splitting Approach for Inconsistency for ALT ............................... 68

Appendix Table 4.22: Side-Splitting Approach for Inconsistency for AST ............................... 68

Appendix Table 4.23: Side-Splitting Approach for Inconsistency for Uric Acid....................... 68

Appendix Table 4.24: Loop-Specific Approach for Inconsistency ............................................ 69

Chapter 5

Table 5.1: Study Exclusion Criteria1.......................................................................................... 118

Table 5.2 Power Calculation for Primary Outcomes Between Water and NSB Arms .............. 119

Table 5.3: Latin Square Randomization for Three Group Crossover Study1 ............................ 120

Table 5.4: Complete List of All Available Study Beverages ..................................................... 120

Table 5.5: Participant Visit Schedule ......................................................................................... 121

Table 5.6 Baseline Characteristics of Randomized Participants1 .............................................. 123

xi

List of Figures

Chapter 2

1Figure 2.1 SSB consumption and the development of obesity, T2DM and Cardiometabolic

Risk[49] ............................................................................................................................................. 9

2Figure 2.2 Absorption, Digestion, Metabolism and Excretion of Acesulfame Potassium,

Saccharin, Aspartame and Sucralose[91] ........................................................................................ 18

3Figure 2.3 Network Meta-Analysis Diagram Combining Direct and Indirect Evidence[141] ....... 26

Chapter 4

Figure 4.1 Literature Search for RCTs of NSBs reporting on Adiposity, Glycemic Control,

Blood Lipids, Blood Pressure, NAFLD and Uric Acid. ............................................................... 54

Figure 4.2 Network Results: Substitution of NSBs for SSBs ...................................................... 55

Figure 4.3: Network Results: Substitution of Water for SSBs .................................................... 56

Figure 4.4 Network Results: Substitution of NSBs for Water ..................................................... 57

Appendix Figure 4.1: Cochrane Risk of Bias Summary for all Included Trials......................... 70

Appendix Figure 4.2: Risk of Bias Proportion for all Included Trials ....................................... 71

Appendix Figure 4.3: Transitivity Analysis_Box Plots Showing the Distribution of the Mean

Age (Years) of the Trials Across the Available Direct Comparisons ........................................... 72

Appendix Figure 4.4: Transitivity Analysis_Box Plots Showing the Distribution of the study

Length (Weeks) of the Trials Across the Available Direct Comparisons .................................... 73

Appendix Figure 4.5: Transitivity Analysis_Box Plots Showing the Distribution of the Sample

Size of the Trials Across the Available Direct Comparisons ....................................................... 74

Appendix Figure 4.6: Transitivity Analysis_Box Plots Showing the Distribution of the % Males

of the Trials Across the Available Direct Comparisons ............................................................... 75

Appendix Figure 4.7: Superplot of the Direct, Indirect and Network Analysis with GRADEc,d

Assessment of the Certainty of the Evidence on the Effect of the Substitution of NSBs for SSBs

on Adiposity Outcomes................................................................................................................. 76



on Glycemic Outcomes ................................................................................................................. 77



on Blood Lipid Outcomes ............................................................................................................. 78

xii



on Blood pressure, NAFLD and Uric Acid Outcomes ................................................................. 79


Assessment of the Certainty of the Evidence on the Effect of the Substitution of Water for SSBs




on Glycemic Outcomes ................................................................................................................. 81








Assessment of the Certainty of the Evidence on the Effect of the Substitution of NSBs for Water




....................................................................................................................................................... 85







Appendix Figure 4.19: Network Diagram for Randomized Controlled Trials Investigating the

Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on Body

Weight ........................................................................................................................................... 88


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on BMI ..... 88


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on Body Fat

% ................................................................................................................................................... 89


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on Waist

Circumference ............................................................................................................................... 89

xiii


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on HbA1c . 90


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on FPG ..... 90


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on 2h-PG .. 91


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on FPI ....... 91


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on HOMA-IR

....................................................................................................................................................... 92


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on LDL-C . 92


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on Non-HDL-

C .................................................................................................................................................... 93


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on TGs ...... 93


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on HDL-C 94

Appendix l Figure 4.32: Network Diagram for Randomized Controlled Trials Investigating the

Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on TC ....... 94


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on SBP ..... 95


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on DBP ..... 95

Appendix Figure 4.35 Network Diagram for Randomized Controlled Trials Investigating the

Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on IHCL ... 96


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on ALT ..... 96


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on AST ..... 97


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on uric acid97

xiv

Chapter 5

Figure 5.1: Study Design1 .......................................................................................................... 117

Figure 5.2: Trial Flow: Screening and Randomization .............................................................. 122

xv

List of Abbreviations

2h-PG = 2h plasma glucose

Ace-K = acesulfame potassium

ADI = Acceptable Daily Intake

ADME = Absorption, Distribution, Metabolism, and Excretion

ALT = alanine transaminase

AST = aspartate transaminase

BMI = body mass index

CHD = coronary heart disease

CI = confidence interval

CIHR = Canadian Institutes of Health Research

CVD = cardiovascular disease

CAC = Codex Alimentarius Commission

CPIR = cephalic phase insulin response

CPR = cephalic phase response

DBP = diastolic blood pressure

DNSG = Diabetes and Nutrition Study Group

EASD = European Association for the Study of Diabetes

FAO = Food and Agriculture Organisation

FPI = fasting plasma insulin

GDM = gestational diabetes

GDP = gross domestic product

GI = gastrointestinal (GI)

GL = glycemic load

GLP-1 = glucagon-like peptide-1

GRADE = Grading of Recommendations Assessment, Development, and Evaluation

HbA1c = hemoglobin A1c

HDL-C = HDL-cholesterol

HOMA-IR = Homeostatic Model Assessment of Insulin Resistance

xvi

iAUC = incremental area under the curve

IDF = International Diabetes Federation

IFG = impaired fasting glucose

IGT = impaired glucose tolerance

IHCL = intra-hepatocellular lipid

JECFA = Joint Expert Committee on Food Additives

LDL-C = LDL-cholesterol

MD = mean difference

MetS = metabolic syndrome

NAFLD = non-alcoholic fatty liver disease

NCEP-ATP III = National Cholesterol and Education Program Adult Treatment Panel III

NNS = non-nutritive sweetener

NOAEL = No Observed Adverse Effect Level

Non-HDL-C = Non-HDL-cholesterol

NSB = non-nutritive sweetened beverage

OGTT = oral glucose tolerance test

PCS = prospective cohort study

RCTs = randomized controlled trials

SBP = systolic blood pressure SBP

SMD = standardized mean difference

SRMA = systematic reviews and meta-analysis

SSB = sugar-sweetened beverage

STOP Sugars NOW = Strategies To OPpose SUGARS with Non-nutritive sweeteners Or Water

STR = sweet taste receptor

T2DM = type 2 diabetes mellitus

TC = total cholesterol

TGs = triglycerides

WHO = World Health Organization

1

Chapter 1: Introduction

Rates of obesity worldwide are increasing and have nearly tripled since 1975. In 2016, over

650 million adults, 18 years of age and older, were living with obesity [1]. These numbers

are concerning as obesity is a well-known risk factor for many debilitating chronic diseases

including type 2 diabetes mellitus (T2DM) [2-4], which has increased in tandem with global

obesity rates [5].

While multiple lifestyle and dietary factors have been implicated in the etiology of obesity

and T2DM, dietary sugars have emerged as the dominant nutrient of concern as their excess

intakes have been associated with weight gain, T2DM and downstream complications

including hypertension and coronary heart disease [6-9]. As a result, the World Health

Organization has called for reductions in free sugars to ≤5-10% of energy [10], with a focus

on sugar-sweetened beverages (SSBs) due to their substantially high free sugar content [5].

This call to action has provided an opportunity for non-nutritive sweetened beverages

(NSBs). Despite the potential benefits of NSBs, recommendations for their consumption are

inconsistent across dietary and clinical practice guidelines [11-18], as concerns exist that they

contribute to an increased risk of the very diseases they were designed to prevent. The

preferred replacement for SSBs in these guidelines is water.

Recent systematic reviews and meta-analyses show an association between NSBs and

increased risk of weight gain, diabetes and cardiometabolic risk [7, 19-23] in prospective

cohort studies, with no effect on these outcomes in the higher quality evidence from

randomized controlled trials (RCTs) [19-21, 24, 25]. The majority of these RCT syntheses,

however, have failed to account for the nature of the comparator, an important determinant in

assessing the effectiveness of NSBs. When non-nutritive sweeteners (NNSs), including

NSBs, displace energy from caloric controls, such as SSBs, the weight of the evidence

indicates benefit for NSBs on several cardiometabolic outcomes [19-21, 24, 26]. As a result,

this thesis will explore the effect of NSBs, as a replacement for SSBs, through 3 prespecified

substitutions: NSBs for SSBs (intended substitution with caloric displacement), water for

2

SSBs (“standard of care” substitution with caloric displacement) and NSBs for water

(matched substitution without caloric displacement) on measures of cardiometabolic risk.

To address this gap in knowledge, this thesis will 1) provide a brief overview of obesity and

diabetes, their relationship to the metabolic syndrome, diagnostic criteria, and management

and prevention; describe how SSBs act as an underlying driver to the development of

cardiometabolic risk factors; SSB consumption trends and recommended intakes; and

summarize the safety, biological pathways and concerns of NNSs on adiposity and glucose

control; 2) report results from a systematic review and network meta-analysis on the effect of

NSBs, as a replacement for SSBs, through 3 prespecified substitutions: NSBs for SSBs,

water for SSBs, and NSBs for water on body weight and cardiometabolic risk; and 3) provide

a rationale, design and baseline characteristics report assessing the effect of substituting

NSBs for SSBs, water for SSBs, and NSBs for water on gut microbiome, glucose tolerance,

and cardiometabolic risk factors: Strategies To OPpose SUGARS with Non-nutritive

sweeteners Or Water trial (STOP Sugars NOW).

3

Chapter 2: Literature Review

2.1 Obesity and Diabetes

2.1.2 Overview of Obesity and Diabetes

Both obesity and diabetes are considered progressive chronic conditions and independent risk

factors for cardiovascular disease [12, 27, 28]. Characterized by abnormal or excessive fat

accumulation, obesity is a major risk factor for type 2 diabetes mellitus (T2DM) [27]. Diabetes is

characterized by the presence of hyperglycemia resulting from impaired and/or defective insulin

secretion and/or action, and represents a group of metabolic disorders that can lead to long-term

microvascular and macrovascular complications [12]. Diabetes can be classified into three main

categories: type 1 diabetes mellitus, T2DM and gestational diabetes (GDM) [29]. Of these types,

the most common is T2DM, representing approximately 90% of people living with diabetes [29].

2.1.3 Prevalence and Economic Burden

The prevalence of obesity and T2DM are increasing. Since 1975 obesity has nearly tripled with

13% of the world’s adult population living with obesity in 2016 at an estimated 650 million [1].

In parallel with the diabetes epidemic is an obesity epidemic and it’s estimated that 80% to 90%

of individuals living with T2DM are overweight or obese [12]. In 2017, an estimated 425 million

adults were living with diabetes and this number is expected to rise to 629 million by 2045 [29].

An aging population coupled with increasing urbanization and economic development are

implicated in the development of these chronic conditions as they are linked with sedentary

lifestyles and poor dietary habits [30]. The economic implications are immense. The global

economic cost of obesity was estimated at USD 2 trillion (2.8% of the gross domestic product

(GDP)) [31], in 2014 and USD 1.3 trillion (1.8% of the GDP) for diabetes in 2017 [32].

2.1.4 Risk Factors and Complications

Risk factors for obesity, although complex, are linked to those that lead to an imbalance of

energy consumed to energy expended, including biological, behavioral, social and environmental

factors [28]. Excess body fat is the number one driver for T2DM, as well as a combination of

genetic and metabolic features, and modifiable lifestyle factors. These include a family history of

the disease, an unhealthy diet, physical inactivity, advancing age, hypertension, cultural origin,

4

impaired glucose tolerance (IGT), smoking, a history of GDM and poor nutrition during

pregnancy [5, 29, 33].

Individuals living with prediabetes, a condition characterized by hyperglycemia below diagnostic

thresholds, are at an increased risk of developing T2DM. Diagnosis of prediabetes is made by the

presence of IGT, impaired fasting glucose (IFG) or a hemoglobin A1c (HbA1c), of 6.0% to

6.4%. IGT occurs when glucose levels are between 7.8-11.0 mmol/L two-hours after an oral

glucose tolerance test (OGTT), whereas IFG is a higher than normal fasting glucose level in the

range of 6.1-6.9 mmol/L [12, 29]. The International Diabetes Federation (IDF) estimates that the

number of people worldwide with IGT was 352.1 million (global prevalence of 7.3%) in 2017

[29].

Included in the risk factors for T2DM is the metabolic syndrome (MetS), a multifaceted and

highly prevalent clustering of abnormalities including central obesity, hypertension, dyslipidemia

and elevated blood glucose [12]. Not only is obesity a significant risk factor in the development

of MetS, but individuals with this syndrome are at significant risk for developing T2DM and

cardiovascular disease (CVD) [34]. A harmonized definition for the diagnosis of the MetS was

established in 2009 and is based on having a minimum of 3 or more criteria (Table 2.1) [35].

1Table 2.1 Harmonized Criterion for the Diagnosis of Metabolic Syndrome[35]

Measure Criteria Cut Points

Men Women

Elevated waist circumference (cm)1

• Canada; USA ≥102 ≥88

• Europids; Middle-Eastern; Sub-Saharan African;

Mediterranean

≥94 ≥80

• Asians; Japanese; South and Central Americans ≥90 ≥80

Elevated TGs (mmol/L) (drug treatment is an alternate indicator2) ≥1.7

Reduced HDL-C (mmol/L) (drug treatment is an alternate

indicator2)

<1.0 <1.3

Elevated BP (mmHg) (antihypertensive drug treatment in an

individual with a history of hypertension is an alternate indicator)

Systolic ≥130 and/or

diastolic ≥85

Elevated FPG (mmol/L) (drug treatment is an alternate indicator3) ≥5.6 BP, blood pressure; FPG, fasting plasma glucose; HDL-C, high-density lipoprotein-cholesterol; T2DM, type 2

diabetes mellitus; TGs, triglycerides 1Population and country specific cut points 2Fibrates and nicotinic acid are commonly used drugs for elevated triglycerides and reduced HDL-C. Individuals

taking one of these drugs are presumed to have high triglycerides and low HDL-C. Omega-3 fatty acids at high

doses indicated high triglycerides

5

3Most individuals with T2DM will have the metabolic syndrome

Aside from T2DM, obesity can lead to the development of several other comorbidities including

coronary artery disease, stroke, osteoarthritis, certain cancers, obstructive sleep apnea and non-

alcoholic fatty liver disease (NAFLD) [36]. T2DM can lead to significant microvascular and

cardiovascular complications. It is one of the leading causes of CVD, blindness, kidney failure,

amputation of the lower-limb and is associated with increased rates of cancer, physical and

cognitive disability, tuberculosis and depression [5, 12]. Eventually, these complications can

result in premature death. The IDF estimated that the number of deaths resulting from diabetes

was 4 million in 2017, accounting for 10.7% of global all-cause mortality among individuals

between 20-79 years [29].

2.1.5 Measures of Overweight and Obesity

The two most common measures for overweight and obesity are through the assessment of body

mass index (BMI) and waist circumference. Taken together, both these measures identify the

level and distribution of adiposity among individuals [36]. As different populations vary in their

level of risk due to differences in fat distributions, the cut-offs for BMI and waist circumference

vary by ethnicity. In general, BMI cut-offs identify individuals at an increased risk for morbidity

and mortality (Table 2.2) [36]. For Asian Americans, lower BMI cut-offs are recommended as

they are at risk for obesity-related diseases below BMI classification values used for non-Asians

[37, 38]. Current population-specific waist circumference cut-offs are the same as those used for

the diagnosis of the MetS as outlined in Table 2.1 [36]. These population-specific waist

circumference values are recommended by the IDF to align the World Health Organization

(WHO) guidelines with the National Cholesterol and Education Program Adult Treatment Panel

III (NCEP-ATP III) guidelines [36, 39-41].

2Table 2.2 BMI Classifications[36]

Classification1 BMI (kg/m²) Disease Risk2

Underweight <18.5 Increased

Normal Weight 18.5-24.9 Least

Overweight 25-29.9 Increased

Obesity ≥30

Class I 30-34.9 High

Class II 35-39.9 Very High

Class II ≥40 Extremely High

6

BMI, body mass index; T2DM, type 2 diabetes mellitus 1For Asian and South Asian individuals BMI classifications are recommended as ≥23 for overweight and ≥25 for

obese [37] 2For T2DM, hypertension and cardiovascular disease relative to normal weight and waist circumference [36]

Even though BMI is often used as the conventional method for assessing adiposity due to its

low-cost and convenience, it cannot distinguish between lean and fat mass or give insight into

the distribution of body fat [42]. Although waist circumference is a better estimate of fat

distribution, particularly visceral fat and its relationship to risk of myocardial infarction (MI) and

CVD, results from INTERHEART’s case-control study of 52 countries by Yusuf et al. [43],

showed that waist-to-hip ratio was a better predictor of both. According to the WHO, waist-to-

hip ratios of ≥0.9cm in men and ≥0.85cm in women substantially increases the risk of metabolic

complications including decreased glucose tolerance, reduced insulin sensitivity and adverse

lipid profiles, all risk factors for T2DM and CVD [44].

Other methods of measuring body fat distribution include skinfold thickness measurements,

bioelectrical impedance, underwater weighing and dual energy x-ray absorptiometry (DXA).

Despite providing more accurate measures of the amount and distribution of body fat, they are

more cumbersome and/or expensive. As a result, BMI, waist circumference and waist-to-hip

ratio are more commonly used [42].

2.1.6 Diagnosis and Targets of Control for Type 2 Diabetes

The three most-common threshold parameters for detecting elevated blood glucose levels are

fasting plasma glucose (FPG), HbA1c, and a 75g 2-hour plasma glucose (2h-PG) OGTT.

According to the 2018 Diabetes Canada guidelines, the diagnostic criteria for T2DM (Table 2.3)

are based on glycemic thresholds associated with microvascular disease, particularly retinopathy,

from venous blood samples and laboratory methods. If one test comes back in the diagnostic

range, the same test should be repeated another day, except in the case of random plasma glucose

where an alternate test should confirm this value among asymptomatic individuals. This

approach results in a confirmatory diagnosis in around 40% to 90% of people presenting with an

initial positive test [12].

7

3Table 2.3 Diagnostic Criterion for Type 2 Diabetes Mellitus [12]

Parameter Diagnostic Criteria for T2DM

FPG (mmol/L) ≥7.0 mmol/L

HbA1c (%) ≥6.5% (adults)

2h-PG in a 75g OGTT (mmol/L) ≥11.1 mmol/L

Random PG (mmol/L) ≥11.1 mmol/L

2h-PG, 2-hour plasma glucose; FPG, fasting plasma glucose; HbA1c, hemoglobin A1c; OGTT, oral glucose

tolerance test; PG, plasma glucose; T2DM, type 2 diabetes mellitus.

Achieving optimal glycemic control is the cornerstone of T2DM management and HbA1c is the

standard test for determining glycated hemoglobin. Evidence indicates that improved glycemic

control (A1C ≤7.0%) reduces the risk of both microvascular and CV complications. To achieve

this target, people living with T2DM should aim for a FPG of 4.7-7.0 mmol/L and a 2h-PG level

of 5.0-10.0 mmol/L [12].

2.1.7 Prevention and Management

Several strategies exist in the prevention and management of obesity and T2DM. The main

preventative public health initiatives are centered around maintaining a healthy body weight,

physical activity and nutrition [12, 28]. Obesity and T2DM management target the adoption of a

healthy lifestyle with a focus on dietary, physical activity, and weight loss approaches, including

behavioural interventions. [12, 28, 29]. Pharmacologic interventions for the treatment of obesity

may be implemented in adults who are unable to attain weight loss through diet and exercise

[28]. Individuals with a BMI ≥40 are candidates for bariatric surgery [36]. For T2DM, the

addition of pharmacotherapy is considered when glycemic control is not achieved within 3

months of behavioral interventions [12]. Given the fiscal costs and personal stigma associated

with these diseases, focusing on prevention will provide substantial economic benefits, while

enhancing quality of life.

2.2 Sugar-Sweetened Beverages

Although several dietary factors are linked to increased body weight and T2DM risk,

considerable evidence indicates an association between excess sugar-sweetened beverage (SSB)

intake and increased risk of obesity, T2DM and cardiometabolic risk [6, 7, 45-48]. The

8

mechanism for these outcomes is likely mediated through an excess intake of energy and through

metabolic effects of certain sugars present in SSBs [49].

Considering that a standard 355mL (12 fl oz) can of regular soda provides ~35-42g of sugar and

140-160 calories, it is likely that SSBs contribute to weight gain by reduced compensation of

subsequent energy intake coupled with a decreased level of satiety [49]. Increased caloric

consumption and weight gain with energy-matched intakes of beverages compared to solid

foods, have been demonstrated in short-term feeding trials suggesting that liquid calories lack

appetite-suppressing characteristics [50-53].

Simultaneously, the glycemic index values of SSBs are moderate-to-high [54], and rapidly spike

postprandial blood glucose and insulin levels [55, 56]. When consumed in large amounts, SSBs

contribute to a high dietary glycemic load (GL), which is of relevance as high-GL diets have

been shown to stimulate insulin resistance [57], aggravate inflammatory biomarkers [58] and are

linked to a higher risk of T2DM and coronary heart disease (CHD) [49].

Furthermore, SSBs contain fructose as a constituent of sucrose or high fructose corn syrup [49].

A moderate consumption of fructose is converted to glucose, lactate and fatty acids in the liver

[49]. However, excess intakes can lead to increased hepatic de novo lipogenesis, triglycerides,

cholesterol, glucose, insulin resistance, uric acid and NAFLD markers [49, 59-63]. A visual

representation of excess SSB consumption and the development of obesity, T2DM and

cardiometabolic risk is shown in Figure 2.1.

9

1Figure 2.1 SSB consumption and the development of obesity, T2DM and Cardiometabolic

Risk[49]

2.2.1 Worldwide Intakes of Sugar-Sweetened Beverages

Despite decreased consumption of SSBs in North America, overall intakes are still above current

WHO [10] recommendations of a maximum of 5-10% of total energy [49, 64-66]. According to

the latest Canadian Community Health Survey Nutrition data, SSBs were the most important

source of free/added sugars among all age groups, where regular soft drinks accounted for 8.9%

of total sugars in 2015 compared to 14.6% in 2004 [64]. Similarly, the most recent results from

the 2011 to 2014 National Health and Nutrition Examination Survey, showed that SSBs provided

6.5% of total calories among adults in the United States [67].

Contrary to high-income countries like Canada and the United States, the increase in

urbanization and beverage marketing in several low and middle-income countries are implicated

in their rising intakes of SSBs [49]. Results from a 2010 survey by Singh et al. [68] of 187

countries found adults living in upper-middle-income and lower-middle-incomes countries had

higher SSB intakes compared to those living in high-income and low-income countries. Of the

21 regions surveyed, the Americas, especially parts of Latin America and the Caribbean, had the

10

highest SSB intakes compared to East Asia which had the lowest [68]. Similarly, in a 2016

review investigating global SSB sales based on calories per person per day, found that per capital

sales of SSBs had increased in lower-middle-income countries and decreased in some high-

income countries [69].

In regard to age and gender, Singh et al. [68] report that adults over the age of 20 consumed an

average of 0.58 (8oz) servings of SSBs per day. Highest SSB intakes were among men aged 20-

39 (1.7 servings per day), whereas women aged 60 plus had the lowest intakes (0.53 servings per

day) [68]. Inequalities in SSB consumption run parallel with inequalities in obesity and T2DM

prevalence in that groups of lower socio-economic status tend to have higher SSB intakes as well

as a higher risk for obesity and T2DM [49].

2.2.2 Guidelines for Sugar-Sweetened Beverage Consumption

Current recommendations set by the WHO and endorsed by national authorities including

Canada, the United States and the United Kingdom, call for reducing free/added sugars to less

than 5-10% of total calories, particularly those coming from SSBs [10, 11, 13, 70]. In line with

these recommendations, public polices exist which aim towards reducing intakes of SSBs

through replacement strategies, taxation, marketing restrictions, limiting availability in schools,

public awareness campaigns and front-of-pack labelling [49]. Of these policies, SSB taxation

shows global momentum as the funds raised can be put toward medical, and public health

initiatives and interventions that target obesity [71]. Evidence from Mexico’s National Institute

of Public Health showed that two years after the Mexican government levied an SSB tax of 1

peso (about $0.07) per liter ($0.02 per ounce) in January 2014, SSB purchases decreased an

average of 7.6% for the general population and by 11.7% among households with the fewest

resources [71, 72].

2.2.3 Replacements for Sugar-Sweetened Beverages in Guidelines

When considering replacement beverages for SSBs, the consensus among several health

authorities is to consume water [11, 13, 14, 73-78]. Non-nutritive sweetened beverages (NSBs)

mimic the sweet taste of SSBs without the associated energy. Despite the potential benefits of

NSBs as displacements of excess calories from SSBs, concerns exist that non-nutritive

sweeteners (NNSs) contribute to an increased risk of obesity and T2DM [19, 21]. Implicated

biological mechanisms for these concerns include impaired sensory and endocrine signaling

11

mediated by the sweet taste receptor [79, 80] and changes to the gut microbiome [80, 81]. As a

result, recommendations for their consumption are inconsistent across dietary and clinical

practice guidelines [11-18]. The US dietary guidelines for Americans and Canada’s Food Guide

discourage the use of NSBs, recommending that water and not NSBs replace SSBs [11, 13]. The

American Heart Association, including diabetes associations in the UK, US, and Canada support

NSBs insofar as they are used to displace calories from sugars and sugar-sweetened beverages

[12, 15, 16, 18]. The European Association for the Study of Diabetes (EASD), on the other hand,

has not made any specific recommendations about NNSs or NSBs [17]. Table 2.4 summarizes

the NNS recommendations in the guidelines from these authoritative bodies.

12

4Table 2.4 Non-Nutritive Sweetener Guideline Recommendations

Guideline Non-Nutritive Sweetener Recommendations

Canada's Food Guide, 2018[11] • “Sugar substitutes do not need to be consumed to reduce intake of free sugars.”

• “No well-established health benefits associated with the intake of sweeteners.”

• “Water, unsweetened milk or fortified soy beverage, and fruit should be offered instead.”

Diabetes Canada, 2018[12] • “The evidence from SRMAs of RCTs,…have shown a weight loss benefit when NNSs are used to

displace excess calories from added sugars (especially from SSBs) in overweight children and

adults without diabetes, a benefit…similar to that seen with… water.”

Dietary Guidelines for

Americans, 2015[13]

• “Replacing added sugars with high-intensity sweeteners may reduce calorie intake in the short-

term, yet questions remain about their effectiveness as a long-term weight management strategy.”

• “…added sugars should be reduced in the diet and not replaced with NNSs, but rather with healthy

options, such as water in place of SSBs.”

American Diabetes

Association, 2020[15]

• “For those who consume SSBs regularly, a low-calorie or NSB may serve as a short-term

replacement strategy, but overall, people are encouraged to decrease both sweetened and NSBs

and use other alternatives, with an emphasis on water intake.”

American Heart Association,

2018[18]

• “Prudent to advise against prolonged consumption of NSBs by children.”

• “For adults who are habitually high consumers of SSBs, …NSBs may be a useful replacement

strategy to reduce intake of SSBs…”

• …“the use of other alternatives to SSBs, with a focus on water (plain, carbonated, and

unsweetened flavored), should be encouraged.”

Diabetes UK, 2018[16] • “For people who are accustomed to sugar sweetened products, NNSs have the potential to reduce

overall energy and carbohydrate intake and may be preferred to sugar when consumed in

moderation”

NSBs, non-nutritive sweetened beverages; NNS, non-nutritive sweetener; SRMA, systematic review and meta-analysis; UK, United Kingdom; WHO, World

Health Organization.

13

2.3 Non-Nutritive Sweeteners

2.3.1 Definition of Non-Nutritive Sweeteners

Non-nutritive sweeteners (NNSs) are sugar substitutes and are also referred to as high-intensity

sweeteners, artificial sweeteners or low-calorie sweeteners [12]. Sugar alcohols, which are also

sugar substitutes, belong to a family of sweeteners called “polyols”. They have a chemical

structure that’s similar to sugar but contain less calories and are only partly absorbed by the

body. As a result, when consumed in amounts greater than 10g, gastrointestinal (GI) discomfort

can occur [82, 83]. NNSs, are significantly sweeter than sucrose (table sugar) and unlike sugar

alcohols their use is not self-limiting from GI symptoms [12, 84]. This distinction is important as

sugar alcohols, aside from erythritol, don’t have Acceptable Daily Intake (ADI) values because

of their potential for GI distress [12]. Even though sugar alcohols are sugar substitutes they

contain more calories than high-intensity sweeteners and, therefore, will not be included in this

thesis when referring to NNSs [85].

2.3.2 Safety and Approval Process

The safety of NNSs has been approved by multiple regulatory bodies, including the European

Food Safety Authority, U.S. Food and Drug Administration and Health Canada [86-88].

Internationally, the Joint Food and Agriculture Organisation (FAO)/WHO Expert Committee on

Food Additives (JECFA) and the Codex Alimentarius Commission (CAC), act to harmonize the

use of NNSs around the world. Comprised of a group of international scientific experts, the

JECFA conducts risk assessments and evaluates the safety of food additives in order to advise

the CAC, FAO, WHO, and its member countries [89] including Canada [90].

NNSs are considered one of the most widely researched food additives and have undergone

extensive toxicology testing in both human and animal safety studies [91]. These toxicology

studies provide the necessary information to identify any potential hazards associated with a

NNS and it’s corresponding No Observed Adverse Effect Level (NOAEL). The NOAEL is the

dose of a substance that does not cause any adverse effects when consumed daily and is

established from long-term, repeated-dose animal studies. Derived from the NOAEL is the ADI

which is calculated by dividing the NOAEL by a typical safety factor of 100. The ADI was

developed by the JECFA and is defined as “the amount of a food additive, expressed on a body

14

weight basis, that can be consumed daily over a lifetime without appreciable health risk”. The

ADI then acts as a safeguard to protect an entire population, including the most susceptible and

those with the highest potential for exposure by ensuring a wide margin of safety [91].

In Canada, NNSs are regulated as food additives under the Food and Drug Regulations and

associated Marketing Authorizations [89]. Only those permitted for use are listed under Health

Canada’s “List of Permitted Sweeteners” [92]. There are currently 9 NNSs (advantame,

acesulfame potassium (Ace-K), aspartame, saccharin, monk fruit extract, neotame, steviol

glucosides, sucralose and thaumatin) approved for use in foods, beverages and/or as table-top

sweeteners by Health Canada [92]. Table 2.5 lists the most common NNSs used in foods,

beverages or as table-top sweeteners in Canada. Of these, aspartame, Ace-K and sucralose are

the most common NNSs found in NSBs [93, 94].

15

5Table 2.5 Common Non-Nutritive Sweeteners Used in Foods and Beverages in Canada[82, 91, 92]

Sweetener Common Forms & Uses Brand Name

Examples

Sweetness

Intensity1

ADI2

Approx. Amount

(mg) in NSBs

No. of NSBs, Tablets or

Packets Equivalent to

ADI3

Acesulfame

Potassium

(Ace-K)

• Although permitted, currently not available

as a table-top sweetener

• Foods and beverages

• Not

applicable

200x 15 42 24 NSBs

Aspartame • Table-top sweetener


• NutraSweet®

Equal®

• Private label

brands4

200x 40

200 14 NSBs

Saccharin • Approved for use in foods and beverages

• Only available in pharmacies as a table-top

sweetener

• Hermesetas® 300x 5

12 (one tablet

Hermesetas®)

28 tablets

Steviol

glycosides

(Stevia)

• Table-top sweetener


• Truvia®

• PureVia®

• Sugar Twin®

Stevia

• Private label

brands4

200-300x 4

>230mg stevia

leaf extract

(Zevia®)5

1 NSB

Sucralose • Table-top sweetener


• Splenda®

• Private label

brandsⱡ

600x 56

18 (Diet

Mountain

Dew®)[95]

19 NSBs

Cyclamate • Table-top sweetener

• Not permitted to be used in foods or

beverages

• Sugar Twin®

• Sweet'N

Low®

• Private label

brands4

30x[96] 11

264 (one packet

Sugar Twin®)

3 packets

ADI, acceptable daily intake; mg, milligram, No., number; NSB, non-nutritive sweetened beverage 1Gram-for-gram sweetness compared to sucrose (table sugar). 2ADI established by the Joint FAO/WHO Expert Committee on Food Additives (JECFA) and reflect those used by Health Canada. Based on mg per kg of body

weight per day. 3Number of NSBs, tablets or packets a 68.2kg (150 pound) person would need to consume to reach the ADI. 4Private label brands include those such as No Name®, President’s Choice®, IrresistiblesTM and Compliments. 5 Received from Zevia customer relations representative. Due to proprietary issues, unable to provide exact quantities of stevia leaf extract per can. 6The ADI for sucralose is 5mg in Canada vs 15mg as per the JECFA.

16

2.3.3 Sweetening Intensity

The sweetening intensity of NNSs are several times higher than sucrose. As a result, very little is

required to elicit a sweet taste. Furthermore, NNSs produce bitter or metallic “off tastes” and are

often found as blends of NNSs to enhance the sweetness intensity and taste profile of a food or

beverage, which further limit the amount of a sweetener that can be used [91]. The sweetening

intensity of NNSs are shown in Table 2.5, along with the number of NSBs, packets or tablets a

68kg (150lbs) person would need to consume in order to reach its respective ADIs.

2.3.4 Non-Nutritive Sweetener Consumption Trends

A 2016 review by Sylvetsky et al. [85] on NNS consumption trends indicated that NSBs are

consumed in the largest proportion of total NNSs globally, exist in many food products, and are

often added to foods and beverages by consumers and product developers. In addition to their

widespread use, the demand for NSBs, is expected to rise [97]. Reasons for the demand are

attributed to increasing awareness and preference for SSB alternatives, and to combat the rise in

health-related problems associated with an excess intake of SSBs [97].

2.3.5 Absorption, Digestion, Metabolism and Excretion of Non-Nutritive

Sweeteners

The biological pathways of NNSs include absorption, distribution, metabolism, and excretion

(ADME), and are established through extensive animal studies in order to determine their safety.

There are considerable differences in the ADME patterns of different NNSs. Knowing the

ADME of each is important in determining if their purported biological effects are indicative of

potential health consequences. A visual representation of the similarities and differences in the

ADME pathways of aspartame, sucralose, Ace-K and saccharin are shown in Figure 2.2.

2.3.5.1 Aspartame

Aspartame is completely digested by digestive enzymes in the GI tract into methanol and the

amino acids phenylalanine and aspartic acid. These byproducts are in the same form as those

derived from foods such as fruits, vegetables and dietary proteins including, meat, fish, poultry,

eggs, diary and legumes. The amino acids are used for protein synthesis and metabolism. Any

excess is excreted in the urine. Methanol is converted to formaldehyde in the liver and used by

17

the body or converted to formic acid and excreted in the urine or broken down to carbon dioxide

and water [91].

2.3.5.2 Sucralose

Sucralose, although similar in structure to sucrose, is not digested into monosaccharides and

undergoes little to no metabolism, thereby providing no calories and no impact on blood glucose.

Only a small amount is absorbed. It is primarily excreted, unchanged, in the feces with

approximately 9-22% excreted in the urine as unchanged sucralose [98]. Due to its low level of

absorption and systemic clearance, the likelihood of sucralose accumulation in the body from

chronic consumption is low [91].

2.3.5.3 Acesulfame Potassium and Saccharin

Ace-K and saccharin are the only two NNSs that are absorbed as intact molecules. Ace-K is

quickly absorbed and distributed to tissues throughout the body via the bloodstream and is

primarily eliminated in the urine with less than 1% excreted in the feces [91].

About 85% to 95% of saccharin is absorbed and excreted in the urine and the remainder is

excreted in the feces [91]. Saccharin was approved for use in foods and beverages in Canada

since 2016, yet is currently only available as the table-top sweetener Hermesetas® [99]. Even

though saccharin is not currently found in beverages in Canada, it’s ADME pathway is worth

mentioning as it has been implicated in the development of glucose intolerance through

alterations in the gut microbiome [81].

18

2Figure 2.2 Absorption, Digestion, Metabolism and Excretion of Acesulfame Potassium,

Saccharin, Aspartame and Sucralose[91]

19

2.3.6 Potential Mechanisms of Non-Nutritive Sweeteners on Body Weight and

Glucose Control

Several potential biological mechanisms suggest an association between NNS intake, weight

gain, glucose intolerance and other negative outcomes. As the ADME pathways vary between

NNSs, so too may their physiological and behavioural effects. These mechanisms, which have

mainly been explored in animal studies, include the activation of sweet taste receptors, changes

in taste preferences, and alterations in the gut microbiome [79, 80, 100].

2.3.6.1 Activation of Sweet Taste Receptors

Sweet taste receptors (STRs), a heterodimeric G-protein coupled receptor (GPCR) comprised of

taste 1 receptor member 2 and taste 1 receptor member 3 subtypes, are activated by sweet tasting

compounds, including NNSs. When activated in the oral cavity, STRs release neurotransmitters

alerting the brain to sweetness [79, 80], while signaling cephalic phase responses (CPRs) which

prime the body to optimize food digestion and nutrient absorption [80]. The most investigated

CPR is the cephalic phase insulin response (CPIR), characterized by a small, neutrally-mediated

release of insulin prior to nutrient absorption [101]. Saccharin is the only NNSs that has been

reported to stimulate the CPIR in humans [102]. Although it’s been postulated that the repeated

activation of the CPIR by saccharin may weaken the body’s ability to respond to sugar

appropriately [103], studies to date in humans indicate non-significant effects on glucose

homeostasis [104].

Aside from the transduction of taste in the mouth, results from animal studies and human cell

lines indicate that the activation of STRs in the pancreas increases insulin secretion [105-107],

and when activated in the intestine, increase the rate of glucose absorption [108-111] and

glucagon-like peptide-1 (GLP-1) [112]. In human studies, when NNSs were administered with

oral glucose, an augmentation of insulin and/or GLP-1 was demonstrated [113-116]. These

hormonal responses, however, were not observed in the majority of human studies that

administered NNSs alone [79].

2.3.6.2 Changes in Taste Preferences and Dietary Intake

As there is an innate preference for sweetness [79, 117] it has been theorized that NNSs, due to

their sweetness intensity, elicit greater preference for sweet tastes leading to poor dietary

choices, increased energy intake, and resultant obesity [79]. This hypothesis has been observed

20

in rodent studies where early-life exposure resulted in a greater preference for sweetness [118].

Overall evidence from adult cross-sectional studies comparing NSB to SSB consumers show an

improvement in diet quality among NSB consumers [117, 119-121]. Furthermore, a 2017 cross-

sectional analysis by Leahy et al. [122] comparing NSB to water consumers showed higher NSB

intakes were associated with a lower intake of carbohydrates, total sugars and added sugars

compared to higher water consumers, suggesting that NSBs, like water, can be a viable option

for reducing sugar intake. Data from intervention trials among adults and children, show that

when sweetness with calories (such as SSBs) is replaced by sweetness without calories (such as

NSBs), energy intake is reduced, and its effect on appetite and energy intake is comparable to

water [20, 117, 123].

2.3.6.3 Gut Microbiome

Several rodent studies have indicated a link between NNSs and alterations of the gut

microbiome, particularly with the NNSs saccharin and sucralose [81, 124-129]. Data on the

effects of NNSs on the human intestinal microbiota, however, are limited and have only been

reported for saccharin, sucralose and steviol glycosides [81, 130, 131].

The most influential human intervention on NNSs and gut microbiome changes was a small,

non-randomized 2014 study by Suez et al. [81]. In this trial, 7 healthy volunteers were given the

max ADI of saccharin (5mg/kg/d) every day for a week and monitored by continuous glucose

measurements and daily OGTTs. Participants that developed poor glycemic responses were

classified as “responders” (n=4), whereas those who did not were classified as “non-responders”

(n=3). Stool from both “responders” and “non-responders” were transferred to germ-free mice.

The mice that received stool from the “responders” developed glucose intolerance suggesting a

causal role for saccharin-induced microbiota alterations. Given the small sample size (n=7), lack

of control group and dose of saccharin given (an amount not found in “real-world” conditions),

the interpretability of this study is severely limited, yet citations for this paper have exceeded

1000 [130].

21

2.3.7 Systematic Reviews and Meta-Analyses of Prospective Cohort Studies

Investigating the Association between Non-Nutritive Sweeteners and Obesity/

Type 2 Diabetes Mellitus Risk

Prospective cohort studies (PCS) consistently demonstrate a positive association for harm with

NNS intake (mainly from NSBs) on weight gain, obesity, T2DM and other cardiometabolic

outcomes including MetS, hypertension, stroke and CVD [7, 19, 20, 22, 132, 133]. Although

PCS provide the highest level of observational evidence to assess the relationship between NSBs

and obesity/T2DM risk, their results may be complicated by reverse causality, residual

confounding or mediation. With reverse causality, many individuals who consume NSBs may be

doing so because they are overweight or obese and, therefore, at an elevated risk for T2DM and

other cardiometabolic risk factors [18, 134]. An emerging method used to address this issue is by

using change in intake to change in outcome analysis [18, 134]. When this type of method is

applied, negative associations are seen between NNS intake and body weight [18, 134-136].

Residual confounding, another inherent limitation with prospective cohort studies, could also

explain the observed association between NNSs and increased risk for T2DM. Even when

relevant covariates such as diet or physical activity are adjusted for, findings may be biased due

to residual confounding [134]. With mediation, an intervening variable or mediator is influenced

by the exposure which in turn influences the outcome [137]. In a 2015 systematic review and

meta-analysis (SRMA) by Imamura et al. [7], the association between a higher intake of NSBs

and risk of T2DM was attenuated from 25% (RR=1.25, 95% CI 1.18 to 1.33) to 8% (RR=1.13,

85% CI 1.02 to 1.15) when adjusted for adiposity, suggestive of mediation for the observed

association as obesity lies on the causal pathway between NSBs and T2DM.

22

2.3.8 Systematic Reviews and Meta-Analyses of Randomized Controlled Trials

Investigating the Effect of Non-Nutritive Sweeteners on Obesity/ Type 2 Diabetes

Mellitus Risk

Contrary to PCSs, RCTs provide the greatest level of protection against bias from confounding

and reverse causality. Results from syntheses of RCTs, however, are not consistent indicating

non-significant or beneficial effects of NNSs on outcomes of adiposity (Table 2.6) and glucose

control (Table 2.7) [19-21, 24, 26, 138]. A major limitation with the majority of these analyses is

the failure to account for the nature of the comparator. When NNSs are compared against energy

matched controls such as water or another NNS, non-significant effects are seen, as these

comparators do not allow for the displacement of energy by the NNS. On the other hand, when

NNSs were compared against sugary controls, including SSBs, a benefit on outcomes of

adiposity and glucose control were observed. Failing to account for the nature of the comparator

resulted in conclusions of “no compelling evidence” for the “intended benefits” of NNSs in a

WHO-commissioned analysis by Toews et al. [21] and a previous SRMA by Azad et al. [19]. In

contrast, a 2020 review by Laviada-Molina et al. [25] and 2016 review by Rogers et al. [20],

distinguished between the effects of NNSs based on the nature of the comparator. In the SRMA

by Rogers et al. [20], although outcomes of glucose control were not assessed, a reduction in

body weight was observed with NNSs compared to sugar-sweetened products among 8 RCTs in

adults (-1.41kg, 95% CI -2.62 to -0.20). When used as a replacement for excess sugars, the

weight of the evidence from SRMAs of RCTs shows benefit with NNS intake on body weight

[20, 21] and fasting glucose [21, 26], with no adverse effect on outcomes of adiposity or glucose

control in individuals with and without diabetes [19-21, 24, 26, 123, 138], a benefit that has

shown to be comparable to displaced excess energy from other interventions, such as water [12].

Other limitations of these reviews include an overly strict selection criteria resulting in the

exclusion of studies with relevant data that could have been included. Azad et al. [19] restricted

studies to ≥6 months, Toews et al. [21] only included studies that specified the NNS, Onakpoya

et al. [26] limited the NNS in their review to stevia-only, while Santos et al. [24] only assessed

aspartame intake, and Wiebe et al. [138] excluded trials with placebo controls. The results from

these reviews were also based on a limited number of studies, thereby reducing their external

validity.

23

6Table 2.6 Systematic Reviews and Meta-Analyses of Randomized Controlled Trials Investigating the Effect of Non-Nutritive

Sweeteners on Adiposity

Study No. of RCTs (participants/population) NNS

Source

Comparator Relevant Results

Laviada-

Molina et

al. 2020

4 (554/ OW, OB adults) NNSs Water • Non-significant effects on BW/BMI with NNSs (SMD=-

0.20, 95% CI -0.62 to -0.23)

13 (1997/ NW, OW, OB children,

adolescents and adults)

NNSs Sugary controls1 • Significant reduction in BW/BMI with NNSs (SMD=-

0.56, 95% CI -0.79 to -0.34)

Toews et

al. 2019

5 (356/ NW, OW, OB adults)

NNS Placebo or sugary

control1

• Non-significant effects on BW with NNSs (MD=1.29kg,

95% CI -2.80 to 0.21)

3 (246/ OW, OB adults)

NNS Sugary control1 • Significant reduction in BW with NNSs (MD=-1.99kg,

95% CI -2.84 to -1.14)


NNS Sugary control1 • Reduced BMI with NNSs (MD=-0.60kg/m2, 95% CI -1.19

to -0.01)

Santos et

al. 2018

3 (179/ OW, OB adults, with or without

T2DM)

NNS

(Asp)

Sugary control1 • Non-significant effects on BW with NNSs (MD=-5.0kg,

95% CI -11.56 to 1.56)

2 (179/ OW, OB adults with or without

T2DM)

NNS

(Asp)

Sugary control1 • Non-significant effects on BW with NNSs (MD=-3.78kg,

95% CI -9.74 to 2.18)

Azad et al.

2017


NSBs, or

mixture

of NNSs

Water or asp avoidance

• Non-significant effects on BW with NNSs (SMD=-0.17,

95% CI -0.54 to 0.21)

3 (242/ NW, OW, OB, or mild HTN

adults)

NSBs, or

stevioside

capsules

Water or placebo • Non-significant effects on BMI with NNSs (MD=-

0.37kg/m2, 95% CI -1.10 to 0.36)

3 (683/ OW, OB adults on weight-loss

program)

NSBs Water

• Non-significant effects on WC with NSBs (SMD=-0.16,

95% CI -0.56 to 0.25)

1 (25/ OW, OB adults on weight-loss

program)

NSBs Water

• Non-significant effects on % body fat with NSBs (MD=-

1.01%, 95% CI -3.01 to 0.99)

Rogers et

al. 2016


NNSs Sugary control1 • Significant reduction in BW with NNSs (MD=-1.41kg,

95% CI -2.62 to -0.20)

1 (641/ NW children) NSBs SSBs

• Significant reduction in BW with NSBs (MD=-1.02kg,

95% CI -1.52 to -0.52)

24

Miller at al.

2014

10 (796/ NW, OW, OB adults) NNS Sugary controls1 • Significant reduction in BW with NNSs (WGMD=-

0.72kg, 95% CI -1.15 to -0.30)

5 (475/ NW, OW, OB adolescents and

adults)

NNS Sugary controls1 • Significantly reduced BMI with NNSs (WGMD=-

0.24kg/m2, 95% CI -0.41 to -0.07)

5 (1000/ NW, OW, OB children,

adolescents and adults)

NNS Sugary controls1 • Significantly reduced fat mass with NNSs (WGMD=-

1.10kg, 95% CI -1.77 to -0.44)

3 (981/ NW, OW, OB children, and

adults)

NSBs SSBs • Significantly reduced WC with NNSs (WGMD=-0.83cm,

95% CI -1.29 to -0.37)

Asp, aspartame; CI, confidence interval; BMI, body mass index; BW, body weight; HTN, hypertension; NNSs, non-nutritive sweeteners; No., number; NSBs,

non-nutritive sweetened beverages; NW, normal weight; OB, obese; OW, overweight; T2DM, type 2 diabetes mellitus; WC, waist circumference; WGMD,

weighted group mean difference. 1Sugary controls include those coming from foods and/or beverages.

25

7Table 2.7 Systematic Reviews and Meta-Analyses of Randomized Controlled Trials Investigating the Effect of Non-Nutritive

Sweeteners on Diabetes Risk

Study No. of RCTs

(participants/population)

NNS Source Comparator Relevant Results

Toews et al.

2019

2 (52/ OW adults)

NNSs Sugary controls1 • Significant reduction in FBG with NNSs (MD = 0.16

mmol/L, 95% CI -0.26 to -0.06)

2 (66/ OW adults)

NNSs Sugary controls1 • No difference observed for plasma insulin (MD = -1.60

pmol/L, 95% CI -8.39 to 5.19)

2 (66/ OW adults)

NNSs Sugary controls1 • No difference observed for HOMA-IR (MD = -0.14, 95% CI

-0.38 to 0.10)

Azad et al.

2017

3 (99/ OW adults)

1 (62/ OW females)

NNSs

NSBs

Water or placebo

Water

• Non-significant effects for insulin resistance (HOMA-IR)

(MD = 0.10, 95% CI, -0.57 to 0.76)

• Non-significant effects on HbA1c (MD = 0.07%, 95% CI -

0.00 to 0.14). Follow-up = 24 weeks

Santos et al.

2018

5 (119/ NW, OW, OB adults with

or without T2DM)

NNS (Asp) Control • Non-significant effects on fasting blood glucose (MD = -

0.03, 95% CI -0.14 to 0.21)

3 (58/ NW, OW, OB adults) NNS (Asp) Sucrose • Non-significant effects on fasting blood glucose (MD = -

0.13, 95% CI -0.95 to 0.69)

4 (101/ NW, OW, OB adults with

or without T2DM)

NNS (Asp) Control • Non-significant effects on insulin levels (MD = -0.13, 95%

CI -0.95 to 0.69)

2 (40/ NW, OW, OB adults)

NNS (Asp) Sucrose • Non-significant effects on insulin levels (MD = -2.54, 95%

CI -11.37 to 6.29)

Wiebe et al.

2011

1 (10/T1DM adults)

NNSs Sugary controls1 • No effect of sweetener type on HbA1c (MD = -0.02%, 95%

CI -0.40 to 0.30) over 4 weeks in a cross-over study

1 (41/OW adults)

NNSs Sugary controls1 • No effect of sweetener type on HOMA-IR (MD = -0.20, 95%

CI -0.58 to 0.18) over 10 weeks in a parallel trial

Onakpoya

et al. 2015

6 (551/ OW, OB adults with or

without T1DM, T2DM, HTN and

hyperlipidemia)

Stevioside

capsules

Matching placebo or

nothing

• Significant reduction in FBG in favour of steviol glycoside

(MD = -0.63 mmol/L, 95% CI -0.90 to -0.36)

Asp, aspartame; BG, blood glucose; CI, confidence interval; FBG, fasting blood glucose; HbA1c, hemoglobin A1c; HOMA-IR, homeostatic model assessment

for insulin resistance; HTN, hypertension; IGT, impaired glucose tolerance; MetS, metabolic syndrome; NNSs, non-nutritive sweeteners; NSBs, non-nutritive

sweetened beverages; NW, normal weight; OB, obese; OW, overweight; T1DM, type 1 diabetes mellitus; T2DM, type 2 diabetes mellitus. 1Sugary controls include those coming from foods and/or beverages

26

2.3.9 Network Meta-Analyses

Compared to standard pairwise meta-analyses, network meta-analyses (NMAs) are extensions

that can be used to simultaneously compare multiple treatments that have not been directly

compared to one another in trials [139]. By comparing multiple treatments in the same model,

NMAs yield more information on the comparative effects of all interventions by using both

direct comparisons of interventions within RCTs and indirect comparisons across trials based on

a common comparator [140]. Figure 2.2 provides a simplistic representation of how direct and

indirect evidence are combined in a NMA [141].

While previous SRMAs have explored NNSs, none to date have simultaneously compared the

effects of NSBs, SSBs and water. Due to the importance of the nature of the comparator in

drawing inferences on the effects of NSBs, NMA techniques provide a statistical means of

leveraging all 3 beverage comparisons that were not directly compared head-to-head in trials.

3Figure 2.3 Network Meta-Analysis Diagram Combining Direct and Indirect Evidence[141]

27

Chapter 3: Rationale and Objectives

3.1 Rationale

Health authorities discourage consumption of SSBs as they have been consistently associated

with weight gain, type 2 diabetes mellitus (T2DM) and downstream complications including

hypertension and coronary heart disease [6-9]. Non-nutritive sweetened beverages (NSBs)

provide a viable alternative for SSBs, yet dietary and clinical practice guidelines are inconsistent

in their recommendations for NSBs out of concern that they do not have established benefits and

recommend that water replace SSBs [11-14, 142]. Systematic reviews and meta-analyses

assessing the totality of evidence from prospective cohort studies on the associations of NSBs on

obesity, diabetes and cardiometabolic risk suffer from inherent limitations of reverse causality,

residual confounding and mediation [18, 134, 137]. Meanwhile, syntheses from RCTs failed to

consider the nature of the comparator by evaluating non-nutritive sweeteners (NNSs) to sugar

(SSBs, sugary foods, etc.) or calorie-matched controls (another NNS, water, placebo, etc.) [19,

21, 24, 26, 138]. There also exists concern that NNSs may induce glucose intolerance through

changes in the gut microbiome [81]. To date, no syntheses of randomized controlled trials

(RCTs) has assessed the effect of substituting NSBs for SSBs, while comparing their

effectiveness to water. Whether NSBs effect cardiometabolic risk factors similar to water in their

intended substitution for SSBs remains unclear.

3.2 Objectives

The overall objective of this thesis is to assess the effect of NSBs as a replacement for SSBs

through 3 prespecified substitutions: NSBs for SSBs (intended substitution with caloric

displacement), water for SSBs (“standard of care” substitution with caloric displacement), and

NSBs for water (matched substitution without caloric displacement) on measures of body

weight, gut microbiome, glucose tolerance and cardiometabolic risk factors.

Specific objectives include:

1. Conduct a systematic review and network meta-analysis of RCTs assessing the effect of

NSBs, as a replacement for SSBs, through 3 prespecified substitutions (NSBs for SSBs,

water for SSBs, and NSBs for water) on measures of body weight and cardiometabolic

28

risk factors using Grading of Recommendations Assessment, Development, and

Evaluation (GRADE) to assess the certainty of evidence.

2. Provide a rationale, design and baseline characteristics report investigating the effect of

NSBs, as a replacement for SSBs, through 3 prespecified substitutions (NSBs for SSBs,

water for SSBs, and NSBs for water), on gut microbiome, glucose tolerance and

cardiometabolic risk factors over 4-weeks in overweight/obese participants who are

regular SSB drinkers: Strategies To OPpose SUGARS with Non-nutritive sweeteners Or

Water trial (STOP Sugars NOW).

29

Chapter 4: Effect of Non-Nutritive Sweetened Beverages as a

Replacement Strategy for Sugar-Sweetened Beverages on Body Weight

and Cardiometabolic Risk: A Network Meta-Analysis of Randomized

Controlled Trials

Néma D McGlynn1,2, Tauseef A Khan1,2, Lily Wang1,3, Roselyn Zhang1,4, Laura Chiavaroli1,2,

Fei Au-Yeung1,2, Elena M Comelli2,5, Vasanti S Malik3,6, James O Hill7, Lawrence A

Leiter1,2,8,9,10, Arnav Agarwal3, Per B Jeppesen11, Dario Rahelić12,13,14, Hana Kahleová15,16, Jordi

Salas-Salvadó17,18,19, Cyril WC Kendall1,2,20, John L Sievenpiper1,2,8,9,10

1Toronto 3D Knowledge Synthesis and Clinical Trials Unit, Clinical Nutrition and Risk Factor Modification Centre,

St. Michael's Hospital, Toronto, ON, Canada; 2Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, ON, Canada; 3Department of Medicine, Faculty of Medicine, University of Toronto, Toronto, ON, Canada; 4Applied Human Nutrition, Mount Saint Vincent University, Halifax, NS, Canada; 5Joannah and Brian Lawson Centre for Child Nutrition, Toronto, ON, Canada; 6Department of Nutrition, Harvard T.H. Chan School of Public Health, Boston, MA, USA; 7Department of Nutrition Sciences, The University of Alabama at Birmingham, Birmingham, Alabama, USA; 8Division of Endocrinology and Metabolism, Department of Medicine, St. Michael’s Hospital, Toronto, ON,

Canada; 9Department of Medicine, Faculty of Medicine, University of Toronto, Toronto, ON, Canada; 10Li Ka Shing Knowledge Institute, St. Michael’s Hospital, Toronto, ON, Canada; 11Department of Clinical Medicine, Aarhus University, Aarhus University Hospital, Aarhus, Denmark; 12Vuk Vrhovac University Clinic for Diabetes, Endocrinology and Metabolic Diseases, Merkur University Hospital,

Zagreb, Croatia; 13University of Zagreb School of Medicine, Zagreb, Croatia; 14University of Osijek School of Medicine, Osijek, Croatia; 15Institute for Clinical and Experimental Medicine, Diabetes Centre, Prague, Czech Republic; 16Physicians Committee for Responsible Medicine, Washington, DC, USA; 17Universitat Rovira i Virgili, Departament de Bioquímica i Biotecnologia, Unitat de Nutrició Humana, Reus, Spain; 18Consorcio CIBER, M.P. Fisiopatología de la Obesidad y Nutrición (CIBERObn), Instituto de Salud Carlos III,

Madrid, Spain; 19Institut d'Investigació Pere Virgili (IISPV), Reus, Spain; 20College of Pharmacy and Nutrition, University of Saskatchewan, Saskatoon, SK, Canada

Corresponding Author: John L Sievenpiper MD, PhD, FRCPC, St. Michael's Hospital, #6138-61

Queen Street East, Toronto, ON, M5C 2T2, CANADA, Tel: 416 867 7475, Fax: 416 867 7495,

email: [email protected]

Keywords: non-nutritive sweetener, low-calorie sweetener, sugar-sweetened beverage, water, cardiometabolic risk,

systematic review, meta-analysis, network meta-analysis

mailto:[email protected]

30

4.1 Abstract

Objective: Concerns exist that non-nutritive sweetened beverages (NSBs) do not have

established benefits, with major dietary guidelines recommending that water and not NSBs

replace sugar-sweetened beverages (SSBs). Whether NSBs improve cardiometabolic risk factors

similar to water in their intended substitution for SSBs is unclear. To inform the update of the

European Association of the Study of Diabetes (EASD) clinical practice guidelines for nutrition

therapy, we conducted a systematic review and network meta-analysis to assess the effect of 3

prespecified substitutions (NSBs for SSBs, water for SSBs and NSBs for water) on body weight

and cardiometabolic risk factors in adults with and without diabetes.

Methods: We searched MEDLINE, Embase and the Cochrane Library through March 2019.

Randomized controlled trials (RCTs) with ≥3week interventions comparing NSBs, SSBs, and/or

water were included. The primary outcome was body weight. Secondary outcomes were other

measures of adiposity, glycemic control, blood lipids, blood pressure, nonalcoholic fatty liver

disease and uric acid. Two independent reviewers extracted data and assessed risk of bias. A

network meta-analysis was performed with data expressed as mean (MD) or standardized mean

(SMD) differences with 95% confidence intervals (CI). GRADE assessed certainty of evidence.

Results: We identified 14 RCTs of 21 trial comparisons, (n=1530) substituting NSBs for SSBs

(7 trials, n=467), water for SSBs (2 trials, n=270) and NSBs for water (7 trials, n=838)

predominantly in adults with overweight/obesity at risk for or with diabetes. Substitution of

NSBs for SSBs reduced body weight (MD,−1.11 kg [95% CI,−1.90, –0.32 kg]), as well as BMI

(−0.32 kg/m2 [−0.58, –0.07 kg/m2]), body fat (−0.60% [−1.03, –0.18%]), triglycerides (−0.24

mmol/L [−0.45, -0.02 mmol/L]), and intrahepatocellular lipid (−0.44 SMD [95% CI, −0.69, –

0.19 SMD]). Substitution of water for SSBs reduced only uric acid (-0.05 mmol/L [-0.08, -0.01

mmol/L]). There was no effect of substituting NSBs for water on any outcome except HbA1c

(+0.21% [+0.02, +0.40%]). The certainty of the evidence was moderate (NSBs for SSBs) and

low (water for SSBs and NSBs for water) for body weight and ranged from low to high for all

other outcomes across all substitutions.

Conclusions: The intended substitution of NSBs for SSBs improves body weight and

cardiometabolic risk factors, showing similar benefits to water and without evidence of harm.

The available evidence supports the use of NSBs as an alternative replacement strategy for SSBs

31

in overweight/obese adults at risk for or with diabetes, over the moderate term. There is a need

for more high-quality RCTs.

Protocol registration: ClinicalTrials.gov identifier, NCT02879500

4.2 Introduction

Sugars have emerged as an important public health concern. The evidence on which this concern

is based derives almost exclusively from sugar-sweetened beverages (SSBs) with excess intakes

of SSBs associated with weight gain and its downstream complications including risk of type 2

diabetes mellitus (T2DM), hypertension, and coronary heart disease (CHD) [6-9]. There is an

urgent need for SSB reduction strategies. Whether non-nutritive sweetened beverages (NSBs) as

a replacement strategy for SSBs have the intended benefits remains unclear. Recent systematic

reviews and meta-analyses, including a WHO-commissioned review [21], have shown NSBs to

be associated with an increased risk of the conditions that they are intended to prevent such as

weight gain, diabetes, and cardiovascular disease (CVD) in prospective cohort studies [19] and

have failed to show the expected weight loss and downstream improvements in cardiometabolic

risk factors in randomized controlled trials (RCTs) [19, 21]. Biological mechanisms involving

impaired sensory and endocrine signaling mediated by the sweet taste receptor [79, 80] and

changes to the microbiome [80, 81] have been offered in support of these observations.

A number of methodological considerations, however, have been raised that limit the inferences

that can be drawn from these data. The available prospective cohort studies are at high risk of

reverse causality (people may consume NSBs because they are already overweight/obese or are

at high risk for weight gain) [130, 134, 143]. Furthermore, the syntheses of RCTs have failed to

account fully for the calories available to be displaced by NSBs with caloric (e.g. SSBs) and

noncaloric (e.g. water, placebo, weight loss diets) comparators pooled together, or noncaloric

comparators used as the sole comparator leading to an underestimation of the effect of NSBs in

their intended substitution for SSBs [130, 134, 143].

The prevailing uncertainties have led to mixed recommendations from authoritative bodies.

Neither the US dietary guidelines for Americans nor Canada’s Food Guide recommend NSBs,

and instead recommend that water and not NSBs replace SSBs [11, 13]. The American Heart

Association (AHA) supports a narrow indication for NSBs, recommending that NSBs replace

SSBs only in adults who are habitual consumers of SSBs with an emphasis on water or

32

unsweetened alternative [18]. Similarly, diabetes associations in the UK, US, and Canada

support NSBs insofar as they are used to displace calories from sugars and SSBs [12, 15, 16].

The European Association for the Study of Diabetes (EASD) has not made any specific

recommendations about non-nutritive sweeteners (NNSs) or NSBs [17]. To update the

recommendations of the EASD, the Diabetes and Nutrition Study Group (DNSG) of the EASD

commissioned a systematic review and meta-analysis to summarize the available evidence from

RCTs of the effect of NSBs, the most important source of NSSs in the diet and a single food

matrix, on intermediate cardiometabolic outcomes. Strength of the evidence was assessed using

the Grading of Recommendations Assessment, Development, and Evaluation (GRADE)

approach [144].

Because of the importance of the comparator in drawing inferences about the effects of NSBs,

we chose to conduct network-meta-analyses as opposed to traditional pairwise meta-analyses to

assess the effect of NSBs through 3 prespecified substitutions: NSBs for SSBs (intended

substitution with caloric displacement), water for SSBs (“standard of care” substitution with

caloric displacement), and NSBs for water (matched substitution without caloric displacement).

4.3 Methods

4.3.1 Design

Our systematic review and network meta-analysis was conducted according to the Cochrane

Handbook for Systematic Reviews of Interventions [145] and reported according to the Preferred

Reporting Items for Systematic Reviews and Meta-Analyses Involving a Network Meta-analysis

(PRISMA-Network Meta-Analysis) [146] (Appendix Table 4.1). The protocol is registered at

ClinicalTrials.gov (identifier, NCT02879500; Results).

4.3.2 Data Sources and Searches

We searched MEDLINE, Embase, and the Cochrane Library from inception through 28 March

2019. The full search strategy is presented in Appendix Table 4.2. Briefly, we searched using

variations of the exposure terms (NSBs, SSBs), outcome terms (adiposity, glycemia, blood

lipids, blood pressure, non-alcoholic fatty liver disease (NAFLD) and uric acid) and study design

terms (randomized controlled trial, randomized, placebo). The search was limited to human

studies with no language restriction. Additionally, reference lists of selected studies and reviews

33

were searched, experts in the field were contacted and Google Scholar searches were conducted

to identify any additional articles.

4.3.3 Study Selection

Appendix Table 4.3 shows the summary of our study selection using the PICOTS framework

[146]. We included RCTs of at least 3-weeks duration that investigated the effect of one of three

beverages (NSBs, SSBs or water) on cardiometabolic risk factors compared to another among

the three beverages. Trials were included if the intervention arm assessed the effect of NSBs,

SSBs or water consumed alone or while on a weight-loss or nutrition education program in

adults. Trials were excluded if they had non-usual intake method (e.g. used NNS capsules that

bypass the oral taste receptors); had a mixed intervention; had a duration of <3 weeks; did not

use a comparator arm of NSBs, SSBs or water; included children, pregnant or breastfeeding

women; or did not provide suitable end-point data. We only studied NNSs in liquid (beverage

trials) as these allow for clear comparisons. Studies of NNSs added to foods, fortified beverages

or nutrient-dense beverages (e.g. milk, juice, etc.) were excluded due to the presence of other

macronutrients. When multiple publications existed for the same study, the article with the most

updated information was included. Published abstracts were not included.

4.3.4 Data Extraction

Two independent reviewers (NM and RZ) assessed the titles and abstracts of all identified

studies and reviewed and extracted relevant data from each report, including study design,

blinding, sample size, participant characteristics, follow-up duration, identification of NSBs,

SSBs and water as an intervention and/or comparator, beverage dosages, outcome data and

funding source. In those trials where the data were included in figures and not provided

numerically, we extracted data using Plot Digitizer V.2.6.8 (http://plotdigitizer.sourceforge.net/).

Additional information was requested from the authors when necessary. Disagreement were

resolved by consensus or where necessary by a third author (TAK).

4.3.5 Outcomes

Outcomes were intermediate measures of cardiometabolic risk as determined by the Clinical

Practice Guidelines Committee of the DNSG of the EASD. The primary outcome was body

weight. Secondary outcomes were other measures of adiposity (body mass index (BMI), % body

34

fat and waist circumference); glycemic control (HbA1c, fasting plasma glucose (FPG), 2h

plasma glucose (2h-PG) during a 75g oral glucose tolerance test (75g-OGTT), fasting plasma

insulin (FPI) and Homeostatic Model Assessment of Insulin Resistance (HOMA-IR)); blood

lipids ((low-density lipoprotein-cholesterol (LDL-C), Non-high-density lipoprotein-cholesterol

(Non-HDL-C), triglycerides (TGs), high-density lipoprotein-cholesterol (HDL-C) and total

cholesterol (TC)); blood pressure (systolic blood pressure (SBP) and diastolic blood pressure

(DBP)); measures of NAFLD (intra-hepatocellular lipid [IHCL], alanine transaminase (ALT)

and aspartate transaminase (AST)); and uric acid. Change-from baseline differences were

preferred over end differences; if not provided, these were calculated from the available data

using published formulas [145].

4.3.6 Risk of Bias Assessment

Risk of bias for each included trial was assessed by the two independent reviewers using the

Cochrane Risk of Bias tool [145]. Assessment was done across 5 domains of bias (sequence

generation, allocation concealment, blinding of participants and personnel, incomplete outcome

data and selective reporting). The risk of bias was assessed as either low (proper methods taken

to reduce bias), high (improper methods creating bias) or unclear (insufficient information

provided to determine the bias level). Difference between reviewers was resolved by consensus.

4.3.7 Data Synthesis

Network meta-analysis, based on a frequentist framework, was conducted using the "network"

suite of commands available in STATA version 15 (College Station, TX: StataCorp LP). The

network meta-analysis synthesized all of the available evidence (direct and indirect effects) and

quantified the pooled network effect of each intervention against every other intervention. We

reported our results as mean differences (MDs) and 95% confidence intervals (CIs). To display

the results for outcomes on the same plot, standardized mean differences (SMDs) were

calculated and pseudo 95% CIs, whereby the SMD 95% CIs were proportionally scaled to the

MD 95% CIs.

We performed a random-effects network meta-analysis for each outcome to compare the three

interventions simultaneously (NSBs, SSBs and water) in a single analysis by combining both

direct and indirect evidence across the selected network of studies. We used change from

baseline values from each study to calculate the mean differences between treatments for each

35

substitution (NSBs for SSBs, water for SSBs and NSBs for water), otherwise we used post-

intervention values. The network diagrams were generated to show the interactions among the

studies included in the network meta-analysis and to illustrate the available direct comparisons

between treatments [147].

Inconsistency was assessed in the direct, indirect, and network estimates. Wassessed interstudy

heterogeneity in the direct and indirect (each pairwise comparison arm) estimates using the

Cochran Q statistic with quantification by the I2 statistic, where an I2≥50%, P<0.10 was

considered an indication of substantial interstudy heterogeneity. We measured incoherence in

the network estimates using both local (loop-specific and side-splitting) and global (design-by-

treatment interaction model) approaches to evaluate the presence of incoherence. The loop-

specific approach looked at the inconsistency in each closed loop in the network [147] while the

side-splitting approach detected direct estimate comparisons that disagreed with the indirect

evidence from the entire network [148]. The design-by-treatment interaction model was applied

as a global approach to simultaneously check for inconsistency from all possible sources in the

network [149]. To explore sources of inconsistency, we conducted sensitivity analyses among

the direct and network comparisons, with the systematic removal of each trial and recalculation

of the pooled estimate. An influential trial was considered a study whose removal changed the

magnitude of the pooled effect by >10% significant. If ≥10 comparisons were available, then we

also conducted a priori subgroup analyses by age, study duration, type of design, disease status,

risk of bias and funding source.

Indirectness was assessed in the indirect comparisons by evaluation of intransitivity (age, study

length, sample size and percentage of males) across the pairwise comparisons comprising the

indirect estimates.

Publication bias was assessed if ≥10 trial comparisons were available, We used comparison-

adjusted funnel plots to assess funnel plot asymmetry [147]

4.3.8 Grading of the Evidence

We assessed the certainty of the evidence using the Grading of Recommendations Assessment,

Development, and Evaluation (GRADE) system [150] with an extension for network meta-

analyses [144] and other recent guidance [151-153] from the GRADE Working group. Evidence

36

was graded as high, moderate, low or very low certainty. Network estimates of RCTs and the

direct and indirect estimates that make-up these network estimates started at high certainty of

evidence and were downgraded by established criteria. Criteria to downgrade included serious

risk of bias (weight of RCTs in the direct estimates) show high risk of bias assessed by the

Cochrane Risk of Bias tool), inconsistency (substantial unexplained heterogeneity [I2≥50%,

P<0.10] in the direct estimates and incoherence in the network estimates by the loop-specific

[p<0.05], side-splitting [p<0.05] and the design-by-treatment interaction model [p<0.10]),

indirectness (presence of factors related to the participants, interventions, and study conditions

that limited the generalizability of the results among the direct estimates and intransitivity in the

indirect estimates), imprecision (the 95% CIs for effect estimates were wide and crossed

prespecified minimally important differences (MIDs) for benefit and/or harm in the direct and

network estimates) and publication bias (evidence of adjusted funnel plot asymmetry).

We used a step-wise approach to apply these downgrades [144], in which the certainty of

evidence from the direct and indirect estimates inform the network estimates. The first step

involved applying the downgrades to the direct estimates. The second step involved applying the

downgrades to the indirect estimates. As the indirect estimates were each comprised of two

pairwise estimates that shared a common comparator, we assessed the certainty of evidence as

the lowest certainty of evidence in the first-order loop. The final step involved grading the

network estimates. If only one of the direct or indirect estimates was available for a given

outcome, then the starting point for the certainty of evidence for the network estimates was based

on that estimate. If both the direct and indirect estimates were available, then the starting point

for the certainty of evidence was based on the estimate that contributed the most weight to the

network estimate. If incoherence and/or imprecision were detected in the network estimates, then

further downgrades were applied for serious inconsistency and serious imprecision, in the

absence of the same downgrades in the direct or indirect estimates used to inform the certainty of

evidence for the network estimate.

4.3.9 Patient Involvement

No patients were directly involved in the development of the research question, selection of the

outcome measures, design and implementation of the study, or interpretation of the results

37

4.4 Results

4.4.1 Search Results

Figure 4.1 shows the flow of the literature. We identified 4541 reports of which 9 met our

eligibility criteria. An additional 5 were found through manual searching. A total of 16 reports of

14 RCTs with 21 trial comparisons (n=1530 participants) were included that assessed the effect

of NSBs for SSBs, water for SSBs, or NSBs for water on body weight, other measures of

adiposity, and cardiometabolic risk [154-169]. Of these, 14 trials reported adiposity (14 for body

weight, 9 for BMI, 5 for % body fat and 6 for waist circumference) [154, 156-162, 164-169], 10

trials reported glycemic control and insulin resistance measurements (10 for FPG, 7 for FPI, and

5 for HOMA-IR, HbA1c and 2hPG [154, 156-162, 164, 168], 8 trials reported blood lipids (7 for

LDL-C and 8 for TGs, HDL-C and TC) [156-162, 164], 5 trials reported blood pressure (4 for

DBP) [156-158, 164, 168], 3 trials reported NAFLD outcomes (2 for IHCL fat, ALT and AST)

[156, 159, 163] and 2 trials reported uric acid concentrations [155, 156].

4.4.2 Available Data

Study and population variables including age, sample size, % females, population description

and intervention details were presented for all RCTs included. The available direct comparisons

between each beverage were represented as a network diagram for each outcome (Appendix

Figures 4.19-4.38).

4.4.3 Trial Characteristics

Table 4.1 shows the key characteristics of all included RCTs, for each beverage comparison, and

the effect on cardiometabolic risk.

The 16 reports of 14 RCTs consisted of 1530 participants (median 72; range 27 to 308) with a

median follow-up of 12 weeks (range 3 to 52). The median age of participants was 34 years

(range 23 to 48). Most participants were overweight or obese with a median BMI of 31 kg/m2

(range 22 to 36). There were more women than men (% men to women, 23:77). Only 8 trials

reported the type of NNS used in the NSBs. Six trials had NSBs with aspartame only. One trial

had NSBs with aspartame and acesulfame potassium (Ace-K) combined and one trial had 4

separate arms with saccharin, rebaudioside A (RebA), sucralose or aspartame as the NNS in the

38

NSBs. The other 6 trials did not report which NNSs were present in the NSBs. The median

beverage doses were 1000 mL/day (range 250 to 2000) for NSBs, 1000 mL/day (range 250 to

1750) for SSBs and 580 mL/day (range 250 to 2000) for water. Nine trials included participants

who were overweight and/or obese. Thirteen trials included healthy individuals, and one with

participants with T2DM. Twelve trials had a parallel design and two were cross-over. Most of

the trials were conducted in Europe (6) and North America (5). Two trials were conducted in the

Middle East and one in South America. Seven trials were funded by agencies (government, not-

for-profit health agency, or university sources), four by industry, and three had agency and

industry funding combined.

4.4.4 Risk of Bias

Appendix Figures 4.1 and 4.2 shows the individual Cochrane Risk of Bias tool assessments for

each of the included trials on body weight and other measures of adiposity, glycemic control,

established blood lipid targets, blood pressure, markers of NAFLD and uric acid. Eight trial

comparisons were rated as unclear risk of bias and 11 were rated as low risk of bias. None were

identified as having a high risk of bias.

4.4.5 Effect of Substitution of NSBs for SSBs

Figure 4.2 shows the network meta-analyses of the effect of the substitution of NSBs for SSBs

on body weight, other measures of adiposity, and cardiometabolic risk factors. The substitution

of NSBs for SSBs reduced the primary outcome body weight (MD= −1.11 kg; 95% CI −1.90, –

0.32 kg) and secondary outcomes BMI (MD= −0.32 kg/m2 ; 95% CI −0.58, –0.07 kg/m2, %

body fat (MD= −0.60%; 95% CI −1.03, –0.18%), TGs [(MD= −0.2 mmol/L; 95% CI −0.45, -

0.02 mmol/L)] and IHCL [(MD= −0.44 SMD; 95% CI −0.69, –0.19 SMD)]. There were no

significant differences for other outcomes.

4.4.6 Effect of Substitution of Water for SSBs

Figure 4.3 shows the network meta-analyses of the effect of the substitution of water for SSBs

on body weight, other measures of adiposity, and cardiometabolic risk factors. The substitution

of water for SSBs did not reduce the primary outcome body weight (MD= −0.09 kg; 95% CI

39

−1.35, 1.17 kg) but did reduce the secondary outcome uric acid (MD= -0.05 mmol/L; 95% CI -

0.08, -0.01 mmol/L). There were no differences in any of the other secondary outcomes.

4.4.7 Effect of Substitution of NSBs for Water

Figure 4.4 shows the network analyses of the effect of the substitution of NSBs for water on

body weight, other measures of adiposity, and cardiometabolic risk factors. The substitution of

NSBs for water did not affect the primary outcome body weight (MD= −1.02 kg; 95% CI −2.15,

0.11 kg) but did increase the secondary outcome HbA1c (MD=0.21%; 95% CI 0.02, 0.40%).

There were no significant differences for any other secondary outcomes.

4.4.8 Inconsistency

Appendix Tables 4.4 to 4.23 shows the side-splitting, and Appendix Table 4.24 shows the

loop-specific assessments of inconsistency (incoherence) in the network estimates. No

significant incoherence was observed by either approach across the 3 substitutions (p<0.05),

except for TGs in the loop-specific assessment (p=0.04). The design-by-treatment model did not

show any significant global inconsistency for any outcome (p<0.10).

4.4.9 Intransitivity (a domain of indirectness, in the indirect estimates)

Appendix Tables 4.3-4.6 shows the evaluation of intransitivity (a domain of indirectness)

among the indirect comparisons by comparing the distribution of the potential effect modifiers

across the available direct comparisons for age, study length, sample size and percentage males.

The assumption of transitivity was met for all indirect comparisons.

4.4.10 Sensitivity Analyses

The systematic removal of each trial among the direct comparisons did not modify the direction,

magnitude or significance of the effect estimates or the evidence of heterogeneity for any of the

outcomes among our 3 prespecified comparisons. We did not conduct sensitivity analyses in the

network as several of the RCTs contained multiple comparison arms in which the removal of one

RCT resulted in the removal of more than one comparison arm.

4.4.11 Subgroup Analyses

As <10 trials were available for any outcome, subgroup analyses was not undertaken.

40

4.4.12 Publication Bias

As <10 trials were available for any outcome, publication bias analyses were not undertaken.

4.4.13 Grading of the Evidence

Appendix Figures 4.7 to 4.18 shows the GRADE assessment for the direct, indirect, and

network meta-analysis. The certainty of the evidence for the substitution of NSBs for SSBs was

moderate for the improvement in the primary outcome body weight owing to a sole downgrade

for imprecision and from low to high for the secondary adiposity and cardiometabolic outcomes

owing to downgrades for inconsistency (waist circumference), indirectness (waist circumference,

HbA1c, 2h-PG, ALT and AST), and imprecision (BMI, waist circumference, HbA1c, 2h-PG,

FPI, LDL-C, Non-HDL-C, TGs, HDL-C, TC, SBP, DBP, IHCL, ALT, AST and uric acid). The

certainty of the evidence for the substitution of water for SSBs was low for the improvement in

the primary outcome body weight owing to a downgrade for inconsistency and imprecision, and

from low to high for the secondary adiposity and cardiometabolic outcomes owing to

downgrades for inconsistency (HbA1c and FPI), indirectness (waist circumference, HbA1c, 2h-

PG, FPI, HOMA-IR, IHCL, ALT, AST and uric acid), and imprecision (BMI, waist

circumference, HbA1c, 2h-PG, FPI, LDL-C, Non-HDL-C, TGs, HDL-C, TC, SBP, DBP, IHCL,

ALT, AST and uric acid). The certainty of the evidence for the substitution of NSBs for water

was low for the improvement in the primary outcome body weight owing to a downgrade for

inconsistency and imprecision, and from low to high for the secondary adiposity and

cardiometabolic outcomes owing to downgrades for inconsistency (waist circumference, HbA1c,

FPI, HOMA-IR, Non-HDL-C and TGs), indirectness, (IHCL, ALT, AST and uric acid), and

imprecision (BMI, waist circumference, HbA1c, 2h-PG, FPI, LDL-C, Non-HDL-C, TGs, TC,

SBP, DBP, IHCL, ALT, AST and uric acid).

4.5 Discussion

The present systematic review and network-meta-analysis of 14 RCTs of 21 trial comparisons

involving 1530 predominantly overweight or obese adult participants who were at risk for or had

diabetes assessed the effect of NSBs on body weight, other measures of adiposity, glycemic

control, blood lipids, blood pressure, markers of NAFLD and uric acid in 3 prespecified

substitutions: NSBs for SSBs (intended substitution with caloric displacement), water for SSBs

41

(“standard of care” substitution with caloric displacement), and NSBs for water (matched

substitution without caloric displacement). The substitution of NSBs for SSBs improved body

weight, BMI, body fat, TGs and IHCL, whereas the substitution of water for SSBs improved

only uric acid. There was no effect of the substitution of NSBs for water with the exception of a

small increase in HbA1c.

4.5.1 Findings in the Context of Existing Studies

We are not aware of any other systematic review and network-meta-analyses that simultaneously

assessed the effects of NSBs, SSBs, and water on intermediate outcomes of cardiometabolic risk.

While several systematic reviews and meta-analyses have assessed the effect of the substitution

of NNSs for caloric comparators in the form of sugary foods/beverages (usually SSBs) either

alone [20] or pooled with substitutions with matched non-caloric comparators such as water,

placebo, or calorie-matched weight loss diets on cardiometabolic risk factors [19-21, 24, 26],

none have quantitatively assessed the effect of NSBs relative to that of water for the intended use

of SSB reduction.

Our findings are in agreement with other systematic reviews and meta-analyses that provided

analyses which allowed for the interpretation of results by comparator. The finding that NSBs

reduce body weight, BMI and body fat in substitution for SSBs with caloric displacement and

have neutral effects on these outcomes in substitution for water without caloric displacement

have been shown by other systematic reviews and meta-analyses of RCTs. Rogers et al. [20]

showed that NNSs reduced body weight in analyses of the substitution of NNSs for a caloric

comparator (sugars in food and beverages) in RCTs involving participants who were

predominantly overweight or obese. Similarly, Toews et al. [21] showed a small improvement in

BMI in analyses of the substitution of NNSs for a caloric comparator (sucrose in food and

beverages) in RCTs involving participants who were predominantly healthy, while Miller et al.

[23] showed these reductions in body weight and BMI, as well as fat mass and waist

circumference in analyses of the substitution of NNSs for the caloric comparator sugars in food

and beverages in RCTs involving participants who were predominantly overweight or obese. A

recent analysis by Laviada-Molina et al. [25] further support our findings where they showed a

significant reduction in unified measurements of body weight and BMI (expressed as SMDs)

with NNSs in substitution for caloric comparators (sugars in foods and beverages) in

42

predominantly overweight or obese participants. In another broader analysis of the effect of

NNSs in substitution for a combination of caloric and non-caloric comparators by Toews et al.

[21] and another analysis that restricted the effect of NSBs in substitution for matched non-

caloric comparators (placebo, water, or weight loss diet) by Azad et al. [19] found no differences

in body weight with NNSs in RCTs involving participants who were predominantly overweight

or obese.

Overall, these findings are consistent with the mechanism that NSBs lead to weight loss through

contributing to a reduction in net energy intake. Although the displacement of calories using

water in substitution for SSBs did not result in the same significant reductions in body weight or

adiposity as seen with NSBs, there were no differences between NSBs and water across these

outcomes, suggesting that lack of differences were likely owing to imprecision.

Downstream improvements in other intermediate cardiometabolic outcomes would be expected

through displacement of calories from SSBs. In addition to weight gain [170], fructose-

containing sugars providing excess calories especially in beverage form have been shown to

increase TGs [59, 60], glucose [61], insulin [61], uric acid [62], and NAFLD markers [63].

Previous systematic reviews and meta-analyses of RCTs suggest that displacement of calories

from sugars using NSBs or water may improve these outcomes. Toews et al. [21] showed that

NNSs in substitution for caloric sugars (sucrose) in food and beverages reduced blood pressure,

being the effect greater when the substitution of NSBs for a non-caloric placebo was accounted

for. We showed that NSBs and water in substitution for SSBs, with the intended displacement of

calories, improve several intermediate cardiometabolic outcomes, including TGs and IHCL for

NSBs, and uric acid for water. Although we failed to show an effect on blood pressure (SBP,

DBP) or other aspects of the lipid profile (LDL-C, non-HDL-C, HDL-C, TC), uric acid, or other

markers of NAFLD (ALT, AST), the direction of the effects favored NSBs or water with 95%

CIs that were imprecise for benefit but excluded meaningful harm based on the prespecified

MIDs. There were also no differences between the two active treatments, NSBs or water, in these

outcomes.

There is a particular concern that NNSs may induce and promote glucose intolerance. Much of

this concern derives from a single intervention study that showed a decrease in 75g-OGTT

derived glucose tolerance after 6 days of saccharin in an NSB format among 7 healthy

43

individuals with subsequent transfer of the phenotype to germ free mice by fecal transplant. This

study, however, had many sources of bias: it saw the effect only with saccharin (an atypical

sweetener not used in NSBs with the exception of TaB); the NNS was given at an extreme dose

(at the maximum ADI of 5mg per kg of body weight/d, given in 3 divided doses per day,

equivalent to 120mg of saccharin or 4 TaB beverages [95]); there was no control group (making

it difficult to disentangle the effect of the trial protocol [6 consecutive days of 75g-OGTTs] from

that of the sweetener); and did a post hoc separation of participants into responders (n=4) and

non-responders (n=3) [81]. It also did not test the question of whether NSBs in the intended

substitution for SSBs would have the same effect. Given the weight loss seen with caloric

displacement, it would be expected that NSBs in substitution for SSBs would improve markers

of glycemic control. This suggestion is supported by the existing systematic reviews and meta-

analyses that allow for interpretation by comparator. Toews et al. [21] showed an improvement

in fasting plasma glucose without a change in fasting insulin or insulin sensitivity (HOMA-IR)

for the substitution of NNSs for caloric sugars (sucrose in food and beverages), whereas Azad et

al. [19] showed no effect on HbA1c or HOMA-IR of the substitution of NSBs for non-caloric

comparators (water or placebo) in overweight, obese or hypertensive participants. We did not see

evidence of benefit or harm across all glycemic control outcomes for caloric displacement with

the substation of either NSBs for SSBs or water for SSBs. We also did not see differences for the

substitution of NSBs for water with the exception of an increase in HbA1c. This difference,

however, is questionable due to imprecision detected in the estimate with two of the three

included direct trials showing improvements in HbA1c for both the water and NSBs arms [161,

162] and the third trial showing no effect of either the water or the NSBs arm, however the

direction of point estimate favored water [159]. More high-quality trials are needed to improve

these estimates.

4.5.2 Strengths and Limitations

This systematic review and network-meta-analysis has several strengths, the first being the use of

network-meta-analysis that unlike traditional pairwise meta-analyses allowed for the assessment

of multiple comparisons by leveraging direct and indirect comparisons with a common

comparator. By using a network meta-analysis it also allowed for the simultaneous assessment of

the 3 prespecified substitutions (NSBs for SSBs, water for SSBs and NSBs for water). Other

strengths include a comprehensive literature search inclusion of only RCTs which provide the

44

greatest protection against bias; no evidence of serious risk of bias among the included trials; and

use of the GRADE approach to assess the certainty of the estimates.

Several of our analyses presented limitations. First, evidence of serious inconsistency was

present on the secondary adiposity outcome of waist circumference in the substitution of NSBs

for SSBs; on the secondary cardiometabolic outcomes of HbA1c and FPI in the substitution of

water for SSBs; and on the primary outcome of body weight and several secondary adiposity

(waist circumference) and cardiometabolic (HbA1c, FPI, HOMA-IR, Non-HDL-C and TGs)

outcomes in the substitution of NSBs for water. We detected substantial heterogeneity (I2≥50%,

P<0.10) in all of these outcomes that could not be explained fully by sensitivity analyses within

the direct pairwise comparisons. We therefore downgraded the evidence for serious

inconsistency.

Second, there was evidence of serious indirectness in several of our analyses. As only one RCT

or less of direct comparisons was available for several secondary adiposity and cardiometabolic

outcomes in the analyses for the substitution of NSBs for SSBs (waist circumference, HbA1c,

2h-PG, ALT and AST), water for SSBs (waist circumference, HbA1c, 2h-PG, FPI, HOMA-IR,

IHCL, ALT, AST and uric acid) and NSBs for water (IHCL, ALT, AST and uric acid), we

downgraded the evidence in all cases for serious indirectness. The moderate median follow-up of

12-weeks was considered another potential source of indirectness across our analyses. Although

there is some uncertainty as to whether the benefits and lack of harm seen with NSBs extend

beyond our 12-week medium follow-up, we felt that any harms would have manifest within 12-

weeks when participants tend to be most adherent. Our analyses also included RCTs with up to 1

year of follow-up that showed no evidence of harm and even benefit. The longest RCT by Peters

et al. [164] of a 12-month intervention of NSBs substituted for water in the context of a 12-

month behavioral weight loss treatment program among participants with overweight and obesity

showed superiority of NSBs for weight loss and weight maintenance. Other large RCTs in

children (not captured in our analyses that were restricted to adults) offer further evidence of

durable benefit. The largest, longest, and highest quality RCT to date which investigated the

double-blinded substitution of NSBs for SSBs among 641 normal-weight children (ages, 4-11

years) over 18 months showed reductions in weight gain and body fat gain [171]. Adherence to

the NSBs in this trial was also confirmed by a urinary biomarker (urinary sucralose). Based on

45

these findings, we chose not to downgrade the evidence for serious indirectness owing to a lack

of long-term follow up and chose instead to make all of our conclusions specific to the moderate

term.

Third, there was evidence of serious imprecision in several of our pooled estimates. The 95%

CIs for the primary outcome of body weight and several secondary adiposity and

cardiometabolic outcomes in the analyses for the substitution of NSBs for SSBs (BMI, waist

circumference, HbA1c, 2h-PG, FPI, LDL-C, Non-HDL-C, TGs, HDL-C, TC, SBP, DBP, IHCL,

ALT, AST and uric acid); water for SSBs (body weight, BMI, waist circumference, HbA1c, 2h-

PG, FPI, LDL-C, Non-HDL-C, TGs, HDL-C, TC, SBP, DBP, IHCL, ALT, AST and uric acid);

and NSBs for water (body weight, BMI, waist circumference, HbA1c, 2h-PG, FPI, LDL-C, Non-

HDL-C, TGs, TC, SBP, DBP, IHCL, ALT, AST and uric acid) crossed our prespecified MIDs.

Accordingly, we downgraded the evidence in all cases for serious imprecision.

Finally, we were also unable to assess for publication bias or conduct subgroup analyses as there

were less than 10 direct RCT comparisons per outcome for each of the 3 prespecified

substitutions. We were also unable to conduct sensitivity analyses in the network as several of

the RCTs contained multiple comparison arms.

Balancing these strengths and limitations, the certainty of the available evidence was assessed as

moderate for the primary outcome of body weight in the substitution of NSBs for SSBs, and low

for the primary outcome of body weight in the substitution of water for SSBs and in the

substitution of NSBs for water. For all other secondary adiposity and cardiometabolic outcomes

the certainty of the evidence ranged from low to high for the 3 prespecified beverage

substitutions. On average, a rating of moderate was given for the substitution of NSBs for SSBs,

and low for the substitution of water for SSBs and NSBs for water. Downgrades were made for

evidence of serious indirectness and serious imprecision in the substitution of NSBs for SSBs

and the substitution of water for SSBs, while downgrades were made for evidence of serious

inconsistency, serious indirectness and serious imprecision in the substitution of NSBs for water.

46

4.5.3 Implications

The findings of this study are highly relevant for informing guidance on the role of NSBs as a

caloric displacement for SSBs in the context of excess sugar intake which has been implicated in

the dual epidemics of obesity and diabetes [172, 173]. Particular focus has been directed to SSBs

as they are the most important source of added/free sugars in several countries worldwide [49,

69, 174] and their overconsumption has been associated with weight gain, diabetes, and

downstream complications of hypertension and coronary heart disease (CHD) [6-9]. These

conditions not only place an enormous impact on the healthcare system but can significantly

reduce an individual’s quality of life [175]. Although the findings from this network meta-

analysis are limited to NSBs, they are of relevance to NNSs as they are added to several foods

and beverages by consumers and product developers in an attempt to reduce calories [85]. Most

importantly, of all product categories containing NNSs as a single food matrix, NSBs are

globally the most consumed source [18, 85].

Water is considered the “standard of care” substitution for SSBs by authoritative bodies [11-13,

15-18]. Yet, many of these health organizations warn against replacing SSBs with NSBs out of

concern that they contribute to an increased risk of obesity and diabetes. A concern that is

echoed in several media headlines[176-184]. Contrary to these concerns, the available evidence

shows that when NSBs are substituted for SSBs, benefits are seen with no evidence of harm on

body weight and cardiometabolic outcomes over the moderate term. For heavy SSB consumers

who are unable to switch to water, NSBs are a viable alternative as they provide the intended

benefit of calorie displacement. This finding is particularly important considering that

individuals struggle to maintain a substantial change to dietary behavior, especially if the change

is associated with cravings or hunger [185]. NSBs, due to their sweet taste, offer an alternative

that may provide the least cognitive bias towards adherence.

4.5.4 Conclusions

When used as the intended substitution for SSBs, NSBs improves body weight and several

markers of cardiometabolic risk factors including adiposity, IHCL and TGs without evidence of

harm. NSBs were shown to be similar to water, the “standard of care” substitution for SSBs, in

their effect on body weight and cardiometabolic risk factors. The available evidence supports the

use of NSBs as an alternative replacement strategy for SSBs in overweight/obese adults at risk

47

for or with diabetes over the moderate term. Due to uncertainty in the estimate, across the 3

prespecified substitutions, further research is needed to improve our estimates. There is a need

for more high quality RCTs of longer duration comparing these beverages to clarify if the

substitution of NSBs for SSBs results in further risk reduction. These RCTs should focus on

quantifying the types of NNS blends in NSBs, while comparing their effectiveness as a caloric

displacement for SSBs against the “standard of care” matched substitution without caloric

displacement of water, to more accurately assess their long-term effects. In this regard, we await

the results of the STOP Sugars NOW trial (ClinicalTrials.gov, NCT03543644); the Effect of

Non-nutritive Sweeteners of High Sugar Sweetened Beverages on Metabolic Health and Gut

Microbiome trial (ClinicalTrials.gov, NCT03259685); the Study of Drinks With Artificial

Sweeteners in People With Type 2 Diabetes trial (ClinicalTrials.gov, NCT03944616); and the

SWITCH trial (ClinicalTrials.gov, NCT02591134). Furthermore, contributions from research

using a range of study designs will likely be required to validate if the intended benefits and lack

of harm with NSBs for SSBs are durable for cardiometabolic risk outcomes. Additional research

should also assess the health impact of our 3 prespecified substitutions among groups not

included, or inadequately captured in this evidence syntheses, such as younger individuals and

higher risk ethnic groups (ex. Aboriginal peoples), to confirm certainty of directionality.

4.5.5 Funding Statement

The Diabetes and Nutrition Study Group (DNSG) of the European Association of the Study of

Diabetes (EASD) commissioned this systematic review and meta-analysis and provided funding

and logistical support for meetings as part of the development of the EASD Clinical Practice

Guidelines for Nutrition Therapy. This work was also supported by the Canadian Institutes of

Health Research (funding reference number, 129920) through the Canada-wide Human Nutrition

Trialists’ Network (NTN). The Diet, Digestive tract, and Disease (3-D) Centre, funded through

the Canada Foundation for Innovation (CFI) and the Ministry of Research and Innovation’s

Ontario Research Fund (ORF), provided the infrastructure for the conduct of this project. NDM

was supported by a CIHR-Masters Award and a Research Training Centre scholarship through

St. Michel’s Hospital. JLS was funded by a PSI Graham Farquharson Knowledge Translation

Fellowship, Diabetes Canada Clinician Scientist award, CIHR INMD/CNS New Investigator

Partnership Prize, and Banting & Best Diabetes Centre Sun Life Financial New Investigator

Award. JSS, gratefully acknowledges the financial support by ICREA under the ICREA

48

Academia program. EMC holds the Lawson Family Chair in Microbiome Nutrition Research at

the University of Toronto. With the exception of the Clinical Practice Guidelines Committee of

the DNSG of the EASD, none of the sponsors had a role in any aspect of the present study,

including design and conduct of the study; collection, management, analysis, and interpretation

of the data; and preparation, review, approval of the manuscript or decision to publish.

4.5.6 Conflict of Interest

NDM was a full-time employee with Loblaws Companies Limited from November 2011 to

December 2017 and in 2019 received the CIHR-Masters Award. TAK has received research

support from the CIHR and an unrestricted travel donation from Bee Maid Honey Ltd. He has

also spoken as an invited speaker at a Calorie Control Council annual general meeting for which

he received an honorarium. LW and RZ report no relevant competing interests. LC is a Mitacs-

Elevate post-doctoral fellow jointly funded by the Government of Canada and the Canadian

Sugar Institute. FAY reports no relevant competing interests. EMC has received research support

from Lallemand Health Solutions and Ocean Spray; and has received consultant fees or speaker

or travel support from Danone, Nestlé and Lallemand Health Solutions. VSM has been on a pro

bono retainer for expert support for the Center for Science in the Public Interest in litigation

related to sugar sweetened beverages and has served as a consultant for the City of San Francisco

for a case related to health warning labels on soda. JOH receives research funding from NIH and

from the National Cattlemen's Beef Association. He is a member of the Scientific Advisory

Committee for General Mills, McCormick Science Institute, and Milk Producers Educational

Program. He is a member of the Board of Trustees for the International Life Sciences Institute

(ILSI). He has equity in Gelesis and Shakabuku, LLC. LAL reports no relevant competing

interests. AA reports no relevant competing interests. PBJ is running a nonprofit public funded

research project entitle: Innosweet- Integrated perception, psychology, and physiology for

maintaining sweetness perception via sugar replacement and reduction for value added healthy

beverage applications (6150-00037A). He is a honorary member of the EUSTAS and was until

2018 Board Member of the Diabetes and Nutrition Study Group (DNSG) of the EASD. DR is

director of Vuk Vrhovac University Clinic for Diabetes, Endocrinology and Metabolic Diseases

at Merkur University Hospital, Zagreb, Croatia. He is the president of Croatian Society for

Diabetes and Metabolic Disorders of Croatian Medical Association. He serves as a Executive

Committee member of Croatian Endocrine Society, Croatian Sociaty for Obesity and Croatian

49

Society for Endocrine Oncology. He was a board member and secretary of IDF Europe and

currently he is the chair of IDF YLD Programme. He has served as a Executive Committee

member of Diabetes and Nutrition Study Group of EASD and currently he serves as a a

Executive Committee member of Diabetes and Cardiovascular Disease Study Group of EASD.

He has served as principal investigator or co-investigator in clinical trials of AstraZeneca, Eli

Lilly, MSD, Novo Nordisk, Sanofi Aventis, Solvay and Trophos. He has received travel support,

speaker fees and honoraria from advisory board engagements and/or consulting fees from

Abbott, Amgen, AstraZeneca, Bayer, Belupo, Boehringer Ingelheim, Eli Lilly, Lifescan –

Johnson & Johnson, International Sweeteners Association, Krka, Medtronic, Mediligo, Novartis,

Novo Nordisk, MSD, Merck Sharp & Dohme, Pfizer, Pliva, Roche, Salvus, Sandoz, Sanofi

Aventis and Takeda. HK is Director of Clinical Research at the Physicians Committee for

Responsible Medicine, a nonprofit organization providing nutrition education and research. JSS

reports serving on the board of and receiving grant support through his institution from the

International Nut and Dried Fruit Council, and Eroski Foundation. Reports serving in the

Executive Committee of the Instituto Danone Spain. Has received research support from the

Instituto de Salud Carlos III, Spain; Ministerio de Educación y Ciencia, Spain; Departament de

Salut Pública de la Generalitat de Catalunya, Catalonia, Spain; European Commission and USA

National Institutes of Health. Has received research support from California Walnut

Commission, Sacramento CA, USA; Almond Board of California, USA; Patrimonio Comunal

Olivarero, Spain; La Morella Nuts, Spain; and Borges S.A., Spain. Reports receiving consulting

fees or travel expenses from Danone; California Walnut Commission, Eroski Foundation,

Instituto Danone - Spain, Nuts for Life, Australian Nut Industry Council, Nestlé, Abbot

Laboratories, and Font Vella Lanjarón. He is on the Clinical Practice Guidelines Expert

Committee of the European Association for the study of Diabetes (EASD), and served in the

Scientific Committee of the Spanisch Food and Safety Agency, and the Spanish Federation of

the Scientific Societies of Food, Nutrition and Dietetics. He is a member of the International

Carbohydrate Quality Consortium (ICQC), and Executive Board Member of the Diabetes and

Nutrition Study Group (DNSG) of the EASD. CWCK has received grants or research support

from the Advanced Food Materials Network, Agriculture and Agri-Foods Canada (AAFC),

Almond Board of California, American Peanut Council, Barilla, Canadian Institutes of Health

Research (CIHR), Canola Council of Canada, International Nut and Dried Fruit Council,

50

International Tree Nut Council Research and Education Foundation, Loblaw Brands Ltd, Pulse

Canada and Unilever. He has received in-kind research support from the Almond Board of

California, American Peanut Council, Barilla, California Walnut Commission, Kellogg Canada,

Loblaw Companies, Quaker (PepsiCo), Primo, Unico, Unilever, WhiteWave Foods/Danone. He

has received travel support and/or honoraria from the American Peanut Council, Barilla,

California Walnut Commission, Canola Council of Canada, General Mills, International Nut and

Dried Fruit Council, International Pasta Organization, Loblaw Brands Ltd, Nutrition Foundation

of Italy, Oldways Preservation Trust, Paramount Farms, Peanut Institute, Pulse Canada, Sun-

Maid, Tate & Lyle, Unilever and White Wave Foods/Danone. He has served on the scientific

advisory board for the International Tree Nut Council, International Pasta Organization,

McCormick Science Institute and Oldways Preservation Trust. He is a member of the

International Carbohydrate Quality Consortium (ICQC), Executive Board Member of the

Diabetes and Nutrition Study Group (DNSG) of the European Association for the Study of

Diabetes (EASD), is on the Clinical Practice Guidelines Expert Committee for Nutrition Therapy

of the EASD and is a Director of the Toronto 3D Knowledge Synthesis and Clinical Trials

foundation. JLS has received research support from the Canadian Foundation for Innovation,

Ontario Research Fund, Province of Ontario Ministry of Research and Innovation and Science,

Canadian Institutes of health Research (CIHR), Diabetes Canada, PSI Foundation, Banting and

Best Diabetes Centre (BBDC), American Society for Nutrition (ASN), INC International Nut

and Dried Fruit Council Foundation, National Dried Fruit Trade Association, The Tate and Lyle

Nutritional Research Fund at the University of Toronto, The Glycemic Control and

Cardiovascular Disease in Type 2 Diabetes Fund at the University of Toronto (a fund established

by the Alberta Pulse Growers), and the Nutrition Trialists Fund at the University of Toronto (a

fund established by the Calorie Control Council). He has received in-kind food donations to

support a randomized controlled trial from the Almond Board of California, California Walnut

Commission, American Peanut Council, Barilla, Unilever, Unico/Primo, Loblaw Companies,

Quaker (Pepsico), Kellogg Canada, and WhiteWave Foods. He has received travel support,

speaker fees and/or honoraria from Diabetes Canada, Mott’s LLP, Dairy Farmers of Canada,

FoodMinds LLC, PepsiCo, The Ginger Network LLC, International Sweeteners Association,

Nestlé, Pulse Canada, Canadian Society for Endocrinology and Metabolism (CSEM), GI

Foundation, Abbott, Biofortis, ASN, Health Sciences North, INC Nutrition Research &

51

Education Foundation, and Physicians Committee for Responsible Medicine. He has or has had

ad hoc consulting arrangements with Perkins Coie LLP, Tate & Lyle, and Wirtschaftliche

Vereinigung Zucker e.V. He is a member of the European Fruit Juice Association Scientific

Expert Panel. He is on the Clinical Practice Guidelines Expert Committees of Diabetes Canada,

European Association for the study of Diabetes (EASD), Canadian Cardiovascular Society

(CCS), and Obesity Canada. He serves as an unpaid scientific advisor for the Food, Nutrition,

and Safety Program (FNSP) and the Technical Committee on Carbohydrates of the International

Life Science Institute (ILSI) North America. He is a member of the International Carbohydrate

Quality Consortium (ICQC), Executive Board Member of the Diabetes and Nutrition Study

Group (DNSG) of the EASD, and Director of the Toronto 3D Knowledge Synthesis and Clinical

Trials foundation. His wife is a former employee of Unilever Canada.

4.5.7 Ethics Approval

Not required

4.5.8 Data Sharing

No additional data are available.

The lead author affirms that this manuscript is an honest, accurate, and transparent account of the

study being reported; that no important aspects of the study have been omitted; and that any

discrepancies from the study as planned (and, if relevant, registered) have been explained.

This is an Open Access article distributed in accordance with the Creative Commons Attribution

Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt,

build upon this work non-commercially, and license their derivative works on different terms,

provided the original work is properly cited and the use is non-commercial. See:

http://creativecommons.org/licenses/by-nc/4.0/.

52

1Table 4.1: Trial Characteristics

Study Subjects Population Age

Mean

(SD)

Sex No.

(%) M:F

NNS Beverage Dose (mL/d)

Setting Design Duration

(Weeks)

Funding

Source

NSB Water SSB

Bonnet et

al. 2018

50 OW/NW, otherwise healthy,

non/low-NNS consumers

31.1

(10.3)

22:28 Asp/

Ace-K

660 660 NA France C 12 A/I

Bruun et

al. 2015

35 OW/OB, otherwise healthy 39

(1.1)

(44:56) Asp 1000 1000 1000 Denmark P 26 A/I

Campos

et al.

2015

27 OW/OB, otherwise healthy,

regular SSB consumers

NR 14:21 NR 1300 NA 1300 Switzerland P 12 A

Engel et

al. 2017

45

OW/OB, otherwise healthy 38.6

(7.6)

(40:60) Asp 1000 1000 1000 Denmark P 26 A/I

Hernánd

ez-

Cordero

et al.

2014

240

OW/OB, otherwise healthy,


33.3

(6.7)

14:13 NR NA 250+ 250+ Mexico P 39 I

Higgins

et al.

2018

93 NW, healthy, non/low-NNS

consumers

22.9

(1.0)

(52:48) Asp 500 500 NA USA P 12 I

Higgins

& Mattes

2019

154


non/low-NNS consumers

27.3

(9.6)

16:29 Sac/ Asp/

RebA/

Suc

1250-

1750

NA 1250-

1750

USA P 12 A

Madjd et

al. 2015

62


regular NSB consumers

32

(6.9)

(36:64) NR 250+ 250+ NA Iran P 24 A

Madjd et

al. 2017

81

OB, T2DM (only on

metformin to control

diabetes), regular NSB

consumers

34.8

(7.2)

0:240 NR 250+ 250+ NA Iran P 24 A

Maersk

et al.

2012

35

OW/OB, otherwise healthy 39

(26)

(0:100) Asp 1000 1000 1000 Denmark P 26 A/I

Peters et

al. 2016


weight stable, regular NSB

consumers

47.8

(10.5)

43:50 NR 710 710 NA USA P 52 I

53

A, agency; Asp, aspartame; Ace-K, acesulfame potassium; F, females; I, industry; M, males; NA, not applicable; No., number; NNS, non-nutritive sweetener;

NSB, non-nutritive sweetened beverage; NW, normal weight; NR, not reported; OB, obese; OW, overweight; RebA, rebaudioside A; Sac, saccharin; SD, standard

deviation; Suc, sucralose; SSB, sugar-sweetened beverage; T2DM, type 2 diabetes mellitus.

*Secondary analyses to Maersk et al. 2012. As more outcomes were reported in the Engel et al. 2017 analysis, data from that trial was used for the majority of

outcomes, except for uric acid (Bruun et al.) and IHCL (Maersk et al).

Reid et

al. 2007

133

NW, weight watchers and

non-weight watchers

31.8

(9.1)

(46:54) Asp 1000 NA 1000 England P 4 A

Reid et

al. 2010

53

OW, otherwise healthy 33.7

(9.9)

67:87 Asp 1000 NA 1000 Scotland P 4 A

Reid et

al. 2014

41

OB, otherwise healthy 35

(9.1)

(44:56) Asp 1000 NA 1000 Scotland P 4 A/I

Tate et

al. 2012



42

(10.7)

0:62 NR 1420-

2000

1420-

2000

NA USA P 26 I

Tordoff

et al.

1990

30

NW, healthy 25.6

(5.3)

(0:100) Asp 1135 NA 1135 USA C 3 A

54

All reports identified through database searching: 4541

(through 28 March, 2019)

MEDLINE: 1052

EMBASE: 1475

The Cochrane Library: 2014

Manual Searches: 5

Total reports after duplicates removed:

3142

Reports excluded by title and abstract: 2778

Non-Human: 137

Cross-Sectional: 19

Case Study: 3

Case-Control: 8

Prospective Cohort: 13

Observational: 81

Commentaries/Editorials/Letters: 5

Review Papers/Conference

Highlights/Guidelines/Protocol Papers: 301

Meta-Analysis: 44

Systematic Review: 48

Drug Study: 282

Inadequate Intervention: 1774

Unsuitable End Points: 63

Reports assessed for full review:

364

Reports excluded by full review: 355

Cross-Sectional: 8

Prospective Cohort: 1

Observational: 11

Review Papers/Conference

Highlights/Guidelines/Protocol Papers: 60

Systematic Review: 3

Drug Study: 2

Inadequate Intervention: 174

Inadequate Comparator: 22

Unsuitable End Points: 41

RCT Acute (<3 weeks): 31

Irretrievable: 2

Total reports included: 16 (14 RCTs)

Trials comparing NSBs vs water: 7

Trials comparing NSBs vs SSBs: 7

Trials comparing SSBs vs water: 2

Adiposity: 14

Glycemic control: 10

Blood lipids: 8

Blood pressure: 5

NAFLD: 3

Uric acid: 2

1Figure 4.1 Literature Search for RCTs of NSBs reporting on Adiposity, Glycemic Control, Blood Lipids,

Blood Pressure, NAFLD and Uric Acid.

55

2Figure 4.2 Network Results: Substitution of NSBs for SSBs

Network analyses of randomized controlled trials investigating the effect of the substitution of NSBs for SSBs on

outcomes of body weight, other measures of adiposity, glycemic control, blood lipids, blood pressure, non-alcoholic

fatty liver disease (NAFLD) and uric acid. Data was pooled using network random effects models and expressed as

mean differences (MD) and 95% confidence intervals (CIs). To display the results for outcomes on the same plot,

standardized mean differences (SMDs) and pseudo 95% CIs (proportionally scaled to the 95% CIs of the MDs) were

calculated. SMDs are represented by a red diamond and 95% CIs by the black horizontal line. 2h-PG, two-hour post

prandial glucose; ALT, alanine transaminase; AST, aspartate transaminase; BMI, body mass index; DBP, Diastolic

Blood Pressure; FPG; Fasting Plasma Glucose; FPI, Fasting Plasma Insulin; HbA1c; hemoglobin A1c; HDL-C,

high-density lipoprotein-cholesterol; HOMA-IR, Homeostatic Model Assessment of Insulin Resistance; IHCL,

intra-hepatocellular lipid; LDL-C, low-density lipoprotein-cholesterol; Non-HDL-C, non-high-density lipoprotein-

cholesterol; NSBs, non-nutritive sweetened beverages; SBP, Systolic Blood Pressure; SSBs, sugar-sweetened

beverages; TC, total cholesterol; TGs, triglycerides; WC, Waist Circumference.

56

3Figure 4.3: Network Results: Substitution of Water for SSBs

Network analyses of randomized controlled trials investigating the effect of the substitution of water for SSBs on

outcomes of body weight, other measures of adiposity, glycemic control, blood lipids, blood pressure, non-alcoholic

fatty liver disease (NAFLD) and uric acid. Data was pooled using network random effects models and expressed as

mean differences (MD) and 95% confidence intervals (CIs). To display the results for outcomes on the same plot,

standardized mean differences (SMDs) and pseudo 95% CIs (proportionally scaled to the 95% CIs of the MDs) were

calculated. SMDs are represented by a red diamond and 95% CIs by the black horizontal line. 2h-PG, two-hour post

prandial glucose; ALT, alanine transaminase; AST, aspartate transaminase; BMI, body mass index; DBP, Diastolic

Blood Pressure; FPG; Fasting Plasma Glucose; FPI, Fasting Plasma Insulin; HbA1c; hemoglobin A1c; HDL-C,

high-density lipoprotein-cholesterol; HOMA-IR, Homeostatic Model Assessment of Insulin Resistance; IHCL,

intra-hepatocellular lipid; LDL-C, low-density lipoprotein-cholesterol; Non-HDL-C, non-high-density lipoprotein-

cholesterol; SBP, Systolic Blood Pressure; SSBs, sugar-sweetened beverages; TC, total cholesterol; TGs,

triglycerides; WC, Waist Circumference.

57

4Figure 4.4 Network Results: Substitution of NSBs for Water

Network analyses of randomized controlled trials investigating the effect of the substitution of NSBs for water

on outcomes of body weight, other measures of adiposity, glycemic control, blood lipids, blood pressure, non-

alcoholic fatty liver disease (NAFLD) and uric acid. Data was pooled using network random effects models

and expressed as mean differences (MD) and 95% confidence intervals (CIs). To display the results for

outcomes on the same plot, standardized mean differences (SMDs) and pseudo 95% CIs (proportionally

scaled to the 95% CIs of the MDs) were calculated. SMDs are represented by a red diamond and 95% CIs by

the black horizontal line. 2h-PG, two-hour post prandial glucose; ALT, alanine transaminase; AST, aspartate

transaminase; BMI, body mass index; DBP, Diastolic Blood Pressure; FPG; Fasting Plasma Glucose; FPI,

Fasting Plasma Insulin; HbA1c; hemoglobin A1c; HDL-C, high-density lipoprotein-cholesterol; HOMA-IR,

Homeostatic Model Assessment of Insulin Resistance; IHCL, intra-hepatocellular lipid; LDL-C, low-density

lipoprotein-cholesterol; Non-HDL-C, non-high-density lipoprotein-cholesterol; NSBs, non-nutritive sweetened

beverages; SBP, Systolic Blood Pressure; TC, total cholesterol; TGs, triglycerides; WC, Waist Circumference.

58

2Appendix Table 4.1: PRISMA-NMAa Checklist

Section/topic Item

# b* Checklist item† Reported

on page #

TITLE

Title 1 Identify the report as a systematic review incorporating a network

meta-analysis (or related form of meta-analysis).

29

ABSTRACT

Structured

summary

2 Provide a structured summary including, as applicable:

Background: main objectives

Methods: data sources; study eligibility criteria, participants, and

interventions; study appraisal; and synthesis methods, such as network

meta-analysis.

Results: number of studies and participants identified; summary

estimates with corresponding confidence/credible intervals; treatment

rankings may also be discussed. Authors may choose to summarize

pairwise comparisons against a chosen treatment included in their

analyses for brevity.

Discussion/Conclusions: limitations; conclusions and implications of

findings.

Other: primary source of funding; systematic review registration

number with registry name.

30

INTRODUCTION

Rationale 3 Describe the rationale for the review in the context of what is already

known, including mention of why a network meta-analysis has been

conducted.

31-32

Objectives 4 Provide an explicit statement of questions being addressed, with

reference to participants, interventions, comparisons, outcomes, and

study design (PICOS).

31-32

METHODS

Protocol and

registration

5 Indicate whether a review protocol exists and if and where it can be

accessed (e.g., Web address); and, if available, provide registration

information, including registration number.

32

Eligibility

criteria

6 Specify study characteristics (e.g., PICOS, length of follow-up) and

report characteristics (e.g., years considered, language, publication

status) used as criteria for eligibility, giving rationale. Clearly describe

eligible treatments included in the treatment networt and note whether

any have been clustered or merged into the same node (with

justification).

32-33, 63

Information

sources

7 Describe all information sources (e.g., databases with dates of

coverage, contact with study authors to identify additional studies) in

the search and date last searched.

34

Search 8 Present full electronic search strategy for at least one database,

including any limits used, such that it could be repeated.

62

59

Study selection 9 State the process for selecting studies (i.e., screening, eligibility,

included in systematic review, and, if applicable, included in the meta-

analysis).

33, 63

Data collection

process

10 Describe method of data extraction from reports (e.g., piloted forms,

independently, in duplicate) and any processes for obtaining and

confirming data from investigators.

33-34

Data items 11 List and define all variables for which data were sought (e.g., PICOS,

funding sources) and any assumptions and simplifications made.

33

Geometry of

the network

S1 Describe methods used to explore the geometry of the treatment

network under study and potential biases related to it. This should

include how the evidence base has been graphically summarized for

presentation, and what characteristics were compiled and used to

describe the evidence base to readers.

34-35

Risk of bias in

individual

studies

12 Describe methods used for assessing risk of bias of individual studies

(including specification of whether this was done at the study or

outcome level), and how this information is to be used in any data

synthesis.

34

Summary

measures

13 State the principal summary measures (e.g., risk ratio, difference in

means). Also describe the use of additional summary measures

assessed, such as treatment rankings and surface under the cumulative

ranking curve (SUCRA) values, as well as modified approaches used to

present summary findings from meta-analyses.

34

Planned methods

of results

14 Describe the methods of handling data and combining results of studies

for each network meta-analysis. This should include, but not be limited

to:

• Handling of multi-arm trials;

• Selection of variance structure;

• Selection of prior distributions in Bayesian analyses; and

• Assessment of model fit.

34-35

Assessment of

inconsistency

S2 Describe the statistical methods used to evaluate the agreement of

direct and indirect evidence in the treatment network(s) studied.

Describe efforts taken to address its presence when found.

35

Risk of bias

across studies

15 Specify any assessment of risk of bias that may affect the cumulative

evidence (e.g., publication bias, selective reporting within studies).

34

Additional

analyses

16 Describe methods of additional analyses if done, indicating which were

pre-specified. This may include, but not be limited to, the following:

• Sensitivity or subgroup analyses;

• Meta-regression analyses;

• Alternative formulations of the treatment network; and

Use of alternative prior distributions for Bayesian analyses (if

applicable).

35-36

60

RESULTS

Study selection 17 Give numbers of studies screened, assessed for eligibility, and included

in the review, with reasons for exclusions at each stage, ideally with a

flow diagram.

37-38, 52-

53

Presentation of

network

structure

S3 Provide a network graph of the included studies to enable visualization

of the geometry of the treatment network.

88-97

Summary of

network

geometry

S4 Provide a brief overview of characteristics of the treatment network.

This may include commentary on the abundance of trials and

randomized patients for the different interventions and pairwise

comparisons in the network, gaps of evidence in the treatment network,

and potential biases reflected by the network structure.

40, 68-71

Study

characteristics

18 For each study, present characteristics for which data were extracted

(e.g., study size, PICOS, follow-up period) and provide the citations.

52-53

Risk of bias

within studies

19 Present data on risk of bias of each study and, if available, any outcome

level assessment.

38

Results of

individual

studies

20 For all outcomes considered (benefits or harms), present, for each study:

(a) simple summary data for each intervention group (b) effect estimates

and confidence intervals, ideally with a forest plot. Modified approaches

may be needed to deal with information from larger networks.

38-39

Synthesis of

results

21 Present results of each meta-analysis done, including

confidence/credible intervals. In larger networks, authors may focus on

comparisons versus a particular comparator (e.g. placebo or standard

care), with full findings presented in an appendix. League tables and

forest plots may be considered to summarize pairwise comparisons. If

additional summary measures were explored (such as treatment

rankings), these should also be presented.

55-57, 76-

87

Exploration for

inconsistency

S5 Describe results from investigations of inconsistency. This may include

such information as measures of model fit to compare consistency and

inconsistency models, P values from statistical tests, or summary of

inconsistency estimates from different parts of the treatment network.

39, 64-69

Risk of bias

across studies

22 Present results of any assessment of risk of bias across studies for the

evidence base being studied.

70-71

Results of

additional

analyses

23 Give results of additional analyses, if done (e.g., sensitivity or subgroup

analyses, meta-regression analyses, alternative network geometries

studied, alternative choice of prior distributions for Bayesian analyses,

and so forth).

39

DISCUSSION

Summary of

evidence

24 Summarize the main findings including the strength of evidence for each

main outcome; consider their relevance to key groups (e.g., healthcare

providers, users, and policy makers).

40-41

Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at

review-level (e.g., incomplete retrieval of identified research, reporting

bias). Comment on the validity of the assumptions, such as transitivity

and consistency. Comment on any concerns regarding network geometry

(e.g., avoidance of certain comparisons).

44-45

Conclusions 26 Provide a general interpretation of the results in the context of other

evidence, and implications for future research.

41-43, 46-

47

FUNDING

61

Funding 27 Describe sources of funding for the systematic review and other support

(e.g., supply of data); role of funders for the systematic review. This

should also include information regarding whether funding has been

received from manufacturers of treatments in the network and/or

whether some of the authors are content experts with professional

conflicts of interest that could affect use of treatments in the network.

47

a Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for

Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(7): e1000097.

doi:10.1371/journal.pmed1000097. For more information, visit: www.prisma-statement.org. b Hutton B, Salanti G, Caldwell DM, Chaimani A, Schmid CH, Cameron C, Ioannidis JP, Straus S, Thorlund K,

Jansen JP, Mulrow C, Catalá-López F, Gøtzsche PC, Dickersin K, Boutron I, Altman DG, Moher D. The PRISMA

extension statement for reporting of systematic reviews incorporating network meta-analyses of health care

interventions: checklist and explanations. Ann Intern Med. 2015 Jun 2;162(11):777-84.

* Boldface indicates new items to this checklist according to PRISMA extension for reporting network meta-

analyses. † Text in italics indicates wording specific to reporting of network meta-analyses that has been added to guidance

from the PRISMA statement.

PICOS = population, intervention, comparators, outcomes, study design.

http://www.prisma-statement.org/

62

3Appendix Table 4.2: Search Strategy

1 Identical search terms were used in all three databases. For all databases, the original search date was 28 March 2019.

Database MEDLINE, EMBASE, Cochrane1

Search

Terms

1. aspartame.mp. 2. exp Aspartame/

3. neotame.mp.

4. saccharin.mp. 5. exp Saccharin/

6. sucralose.mp.

7. stevia.mp. 8. exp Stevia/

9. acesulfame.mp.

10. exp Sweetening Agents/ 11. sugar substitute*.mp.

12. noncaloric.mp.

13. non-caloric.mp. 14. nonnutritive.mp.

15. non-nutritive.mp.

16. no calorie*.mp. 17. low calorie sweeten*.mp.

18. sugar-free.mp.

19. artificial sweet*.mp. 20. diet beverage.mp.

21. diet soda.mp.

22. exp Carbonated Beverages/ 23. ssb.mp.

24. sugar*sweetened beverage*.mp.

25. Fructose/ 26. fructose.mp.

27. dietary sucrose/

28. sucrose.mp. 29. High Fructose Corn Syrup/

30. High Fructose Corn Syrup.mp.

31. Soft drink*.mp. 32. cola.mp.

33. sugar*.mp.

34. 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 11 or 12 or

13 or 14 or 15 or 16 or 17 or 18 or 20 or 21 or 22 or

23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or

32 or 33

35. Blood Pressure/

36. systolic blood pressure.mp. 37. SBP.mp.

38. diastolic blood pressure.mp.

39. DBP.mp. 40. 35 or 36 or 37 or 38 or 39

41. glyc*m*.mp. 42. Hemoglobin A, Glycosylated/

43. glyc*mia.mp. 44. insulin*.mp.

45. gly*albumin.mp.

46. OGTT.mp.

47. hba1c.mp.

48. HOMA*.mp.

49. fructosamine*.mp. 50. Insulin/

51. exp Glucose/

52. Glucose Tolerance Test/ 53. 41 or 42 or 43 or 44 or 45 or 46 or 47 or 48 or 49 or 50 or

51 or 52

54. triglyceride.mp.

55. triacylglycerol.mp.

56. VLDL.mp. 57. very low density lipoprotein.mp.

58. lipid*.mp.

59. lipids/ 60. cholesterol/

61. cholesterol.mp.

62. lipoprotein.mp. 63. lipoproteins/

64. (hdl or high density lipoprotein).mp.

65. (ldl or low density lipoprotein).mp. 66. exp hyperlipidemias/

67. apolipoprotein*.mp.

68. non-HDL.mp. 69. 54 or 55 or 56 or 57 or 58 or 59 or 60 or 61 or 62 or 63 or

64 or 65 or 66 or 67 or 68

70. fatty liver.mp.

71. non-alcoholic fatty liver disease/

72. NAFLD.mp. 73. transaminases/

74. alanine transaminase/

75. alt.mp. 76. aspartate aminotransferase/

77. ast.mp.

78. IHCL.mp.

79. Intrahepatocellular lipid.mp.

80. transamin*.mp.

81. 70 or 71 or 72 or 73 or 74 or 75 or 76 or 77 or 78 or 79 or

80

82. exp uric acid/ 83. uric acid.mp.

84. urate.mp.

85. hyperuricemia/ 86. hyperuricemia.mp.

87. hyperuricaemia.mp.

88. uric.mp.

89. 82 or 83 or 84 or 85 or 86 or 87 or 88

90. randomized controlled trial.mp,pt.

91. randomized.mp. 92. placebo.mp.

93. 90 or 91 or 92

94. 40 or 53 or 69 or 81 or 89 95. 34 and 93 and 94

Database Total

Medline 1052

Embase 1475

Cochrane 2014

Manual Searches 5

TOTAL 4546

63

4Appendix Table 4.3: PICOTSb Framework

Participants Intervention Comparison Outcome Time Study Design

Adult men and

women

excluding

pregnant or

breastfeeding

women and

children. All

health

backgrounds

NSBs or

SSBs or

water

NSBs or

SSBs or water

Adiposity,

glycemic

control,

established

blood lipid

targets, blood

pressure, non-

alcoholic fatty

liver and uric

acid

≥ 3

weeks

Human

randomized

controlled

trials

bHutton B, Salanti G, Caldwell DM, Chaimani A, Schmid CH, Cameron C, Ioannidis JP, Straus S, Thorlund K,

Jansen JP, Mulrow C, Catalá-López F, Gøtzsche PC, Dickersin K, Boutron I, Altman DG, Moher D. The PRISMA

extension statement for reporting of systematic reviews incorporating network meta-analyses of health care

interventions: checklist and explanations. Ann Intern Med. 2015 Jun 2;162(11):777-84.

64

5Appendix Table 4.4: Side-Splitting Approach for Inconsistency for Body Weight

Side Direct Indirect Difference

Coef. Std. Err. Coef. Std. Err. Coef. Std. Err. P>z

Water

for SSBs

0.4057214 1.177256 0.042085 0.839443 -0.44781 1.443613 0.756

NSBs

for SSBs

1.066648 0.4535466 1.519124 1.39197 -0.45248 1.464935 0.757

NSBs

for

Water

1.100907 0.7015187 0.668898 1.277997 0.432009 1.456246 0.767

6Appendix Table 4.5: Side-Splitting Approach for Inconsistency for BMI

Side Direct

Indirect

Difference


Water for

SSBs

0.1778188 0.282524 -0.7871 0.450022 0.609281 0.529387 0.25

NSBs for

SSBs

0.3601485 0.138045 -0.26236 0.519513 0.622504 0.537329 0.247

NSBs for

Water

0.4060461 0.41773 0.189944 0.319592 -0.59599 0.525259 0.257

BMI, body mass index; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

7Appendix Table 4.6: Side-Splitting Approach for Inconsistency for Body Fat %

Side Direct

Indirect

Difference


Water for

SSBs

-0.192686 0.7165101 -1.147723 1.942673 1.128455 2.067054 0.585

NSBs for

SSBs

0.6185242 0.2192872 0.6197683 2.074374 1.238293 2.085881 0.553

NSBs for

Water

0.5359688 1.898476 0.6054734 0.7514356 -1.141442 2.039145 0.576

NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

8Appendix Table 4.7: Side-Splitting Approach for Inconsistency for Waist Circumference

Side Direct

Indirect

Difference


Water for

SSBs

0.30000

01

2.03127

8

0.820217

2

141.4286 -

0.5202172

141.4432 0.99

7

NSBs for

Water

0.82023

07

1.02721

7

-

0.220309

1

282.9657 1.04054 282.9689 0.99

7


65

9Appendix Table 4.8: Side-Splitting Approach for Inconsistency for HbA1c

Side Direct Direct Indirect

Differenc

e


Water for SSBs

-0.05

0.219310

6

-

0.11329

0.17317

4 0.063291

0.27943

9

0.82

1

NSBs for SSBs

-0.11 0.125894

-

0.17329

0.24947

3 0.063291

0.27943

9

0.82

1

NSBs for

Water -

0.223292

0.118911

1 -0.16

0.25287

7 -0.06329 0.27944

0.82

1 HbA1c, hemoglobin A1c; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

10Appendix Table 4.9: Side-Splitting Approach for Inconsistency for FPG


Difference


Water for

SSBs

-0.03654 0.13924 -0.18162 0.107769 0.145082 0.175492 0.408

NSBs for

SSBs

0.168507 0.092779 0.004328 0.168865 0.164179 0.191957 0.392

NSBs for

Water

-0.00184 0.040021 0.135048 0.200939 -0.13689 0.204885 0.504

FPG, fasting plasma glucose; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

11Appendix Table 4.10: Side-Splitting Approach for Inconsistency for 2h-PG


Difference


NSBs for

SSBs

0.29122 0.378731 1.71E-03 142.5415 -0.29293 142.542 0.998

NSBs for

Water

0.19005 0.099326 -0.7558 284.7283 0.565751 284.7282 0.998

2h-PG, two-hour post prandial glucose; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

12Appendix Table 4.11: Side-Splitting Approach for Inconsistency for FPI


Difference


Water for

SSBs

12.09007 21.76729 -32.2253 15.84592 44.31534 29.52399 0.133

NSBs for

SSBs

12.1009 10.80856 -58.2259 46.91402 70.32684 48.63387 0.148

NSBs for

Water

-7.99187 5.604813 31.18152 42.82504 -39.1734 42.73854 0.359

FPI, fasting plasma insulin; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

66

13Appendix Table 4.12: Side-Splitting Approach for Inconsistency for HOMA-IR


Difference


Water for

SSBs

0.32139 0.496521 -0.86131 0.670215 1.182697 0.915703 0.197

NSBs for

SSBs

0.241025 0.3648 -1.02249 1.101164 1.26352 1.182708 0.285

NSBs for

Water

-0.04597 0.198863 1.566582 1.878476 -1.61255 1.882096 0.392

HOMA-IR, homeostatic model assessment of insulin resistance; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

14Appendix Table 4.13: Side-Splitting Approach for Inconsistency for LDL-C


Difference


Water for

SSBs

0.08227

0.13972 0.03716 0.15838 0.07855 0.20937 0.70800 NSBs for

SSBs 0.06182 0.14030 -0.12531 0.16127 0.18713 0.21375 0.38100 NSBs for

Water 0.02700 0.04922 0.07387 0.22275 -0.04687 0.22814 0.83700 LDL-C, low-density lipoprotein cholesterol, NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

15Appendix Table 4.14: Side-Splitting Approach for Inconsistency for Non-HDL-C


Difference


Water for

SSBs

-0.04854 0.105695 -0.09359 0.150071 0.04505 0.183311 0.806

NSBs for

SSBs

0.158931 0.124875 0.00913 0.124738 0.149801 0.176104 0.395

NSBs for

Water

0.021362 0.067712 0.026765 0.182856 -0.0054 0.196445 0.978

Non-HDL-C, Non-high-density lipoprotein cholesterol, NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

16Appendix Table 4.15: Side-Splitting Approach for Inconsistency for TGs


Difference


Water for

SSBs

-0.28485 0.1587012 -0.0957 0.157552 -0.18915 0.222016 0.394

NSBs for

SSBs

0.217221 0.1340943 0.279237 0.186124 -0.06202 0.226371 0.784

NSBs for

Water

0.056847 0.0561295 -0.10666 0.241273 0.163505 0.248033 0.51

NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages; TGs, triglycerides.

67

17Appendix Table 4.16: Side-Splitting Approach for Inconsistency for HDL-C


Difference


Water for

SSBs

0.040828 0.052674 -0.01915 0.071828 0.059977 0.089072 0.501

NSBs for

SSBs

0.0106076 0.059408 -0.06617 0.063332 0.076774 0.086835 0.377

NSBs for

Water

0.0089119 0.024449 0.043649 0.092413 -0.05256 0.095592 0.582

HDL-C, high-density lipoprotein cholesterol, NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

18Appendix Table 4.17: Side-Splitting Approach for Inconsistency for TC


Difference


Water for

SSBs

-0.0606 0.2349 -0.0937 0.1625 -0.0331 0.2825 0.907

NSBs for

SSBs

0.1603 0.1392 -0.1716 0.2913 0.3319 0.3196 0.299

NSBs for

Water

0.0212 0.0635 -0.0829 0.3428 0.1041 0.3488 0.765

NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages; TC, total cholesterol.

19Appendix Table 4.18: Side-Splitting Approach for Inconsistency for SBP


Difference


Water for

SSBs

0.460598 2.588089 -3.12636 5.446124 3.586955 6.020967 0.551

NSBs for

SSBs

5.694315 3.635833 0.220911 3.470725 5.473405 5.045351 0.278

NSBs for

Water

2.319843 1.840343 7.140843 6.00167 -4.821 6.25626 0.441

NSBs, non-nutritive sweetened beverages; SBP, systolic blood pressure; SSBs, sugar-sweetened beverages.

20Appendix Table 4.19: Side-Splitting Approach for Inconsistency for DBP


Difference


Water for

SSBs

-0.17673 1.634782 -5.322171 3.461488 5.145445 3.813087 0.177

NSBs for

SSBs

5.158686 2.408139 -0.8695577 2.334642 6.028243 3.349012 0.072

NSBs for

Water

0.989036 2.240188 4.606612 4.569729 -3.61758 5.252017 0.491

DBP, diastolic blood pressure; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

68

21Appendix Table 4.20: Side-Splitting Approach for Inconsistency for IHCL


Difference


Water for

SSBs

0.46000 0.21000 0.34051 0.24464 0.11949 0.32241

NSBs for

SSBs

0.42051 0.14089 0.54000 0.29000 -0.11949 0.32241

NSBs for

Water

-0.08000 0.20000 0.03949 0.25288 -0.11949 0.32241

IHCL, intra-hepatocellular lipid; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

22Appendix Table 4.21: Side-Splitting Approach for Inconsistency for ALT


Difference


NSBs for

SSBs

3.5000 7.1000 0.0000 223.6259 3.5000 223.7386 0.988

NSBs for

Water

-1.1998 1.4881 5.8002 447.2031 -7.0000 447.2007 0.988

ALT, alanine transaminase; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

23Appendix Table 4.22: Side-Splitting Approach for Inconsistency for AST


Difference


NSBs for

SSBs

1.5 3.25 2.36E-04 223.6165 1.499764 223.6402 0.995

NSBs for

Water

-0.2 1.781909 2.800086 447.2568 -3.00009 447.2532 0.995

AST, aspartate transaminase; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

24Appendix Table 4.23: Side-Splitting Approach for Inconsistency for Uric Acid


Difference


Water for

SSBs

0.05 0.023463 0.036245 0.038575 0.013756 0.045151

NSBs for

SSBs

0.026253 0.0201564 0.0399999 0.039979 -0.01375 0.044773

NSBs for

Water

0.01 0.0324097 0.0237565 0.031373 -0.01376 0.045107


69

25Appendix Table 4.24: Loop-Specific Approach for Inconsistency

Outcome P-value Inconsistency

Body weight P=0.074 Not present

BMI P=0.8 Not present

Body fat P=0.98 Not present

Waist circumference No triangular loop found

HbA1c Unable to calculate*

FPG P=0.25 Not present

2h-PG No triangular loop found

FPI P=0.75 Not present

HOMA-IR P=0.76 Not present

LDL-C P=0.95 Not present

Non-HDL-C P=0.66 Not present

TGs P=0.04 Present

HDL-C P=0.49 Not present

TC P=0.63 Not present

SBP P=0.15 Not present

DBP P=0.57 Not present

IHCL Unable to calculate*

ALT No triangular loop found

AST No triangular loop found

Uric acid Unable to calculate* 2h-PG, two-hour post prandial glucose; ALT, alanine transaminase; AST, aspartate transaminase; BMI, body mass

index; DBP, diastolic blood pressure; FPG, fasting plasma glucose; FPI, fasting plasma insulin; HbA1c, hemoglobin

A1c; HDL-C, high-density lipoprotein-cholesterol; HOMA-IR, Homeostatic Model Assessment of Insulin

Resistance; IHCL, intra-hepatocellular lipid; LDL-C, low-density lipoprotein-cholesterol; Non-HDL-C, non-high-

density lipoprotein-cholesterol; SBP, systolic blood pressure; TC, total cholesterol; TGs, triglycerides.

* Calculation issues with loop-specific approach in Stata with multiple-arm trials as model did not converge,

therefore, “unable to calculate” for all missing ones.

70

Figures

5Appendix Figure 4.1: Cochrane Risk of Bias Summary for all Included Trials

71

6Appendix Figure 4.2: Risk of Bias Proportion for all Included Trials

72

7Appendix Figure 4.3: Transitivity Analysis_Box Plots Showing the Distribution of the Mean

Age (Years) of the Trials Across the Available Direct Comparisons

73

8Appendix Figure 4.4: Transitivity Analysis_Box Plots Showing the Distribution of the study

Length (Weeks) of the Trials Across the Available Direct Comparisons

74

9Appendix Figure 4.5: Transitivity Analysis_Box Plots Showing the Distribution of the Sample

Size of the Trials Across the Available Direct Comparisons

75

10Appendix Figure 4.6: Transitivity Analysis_Box Plots Showing the Distribution of the % Males

of the Trials Across the Available Direct Comparisons

76

11Appendix Figure 4.7: Superplot of the Direct, Indirect and Network Analysis with GRADEc,d


on Adiposity Outcomes

Network analyses of randomized controlled trials investigating the effect of the substitution of NSBs for SSBs on outcomes of adiposity. Data

was pooled using network random effects models and expressed as mean differences (MD) and 95% confidence intervals (CIs). MDs for the direct pair-wise differences were synthesized with the use of generic inverse variance random effects models. MDs for the indirect pair-wise

differences were synthesized with the use of generic inverse variance random effects models and 95% CIs using a common comparator. To

display the results for outcomes on the same plot, standardized mean differences (SMDs) and pseudo 95% CIs (proportionally scaled to the 95% CIs of the MDs) were calculated. SMDs are represented by a red diamond and 95% CIs by the black horizontal line. Filled, black boxes represent

which categories downgrades were made among the direct, indirect and network estimates. The starting point for the certainty of the evidence for

the network estimate was based on the certainty of the evidence for the direct or indirect estimate that contributed the most weight to the network estimate.

BMI, body mass index; GRADE, Grading of Recommendations, Assessment, Development, and Evaluation; MID, Minimally Important

Difference; NE, not estimable; No., number; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages; WC, waist circumference. 1The certainty of evidence for the indirect estimates is based off the lowest rating in the first order loop. 2We rated down the evidence for serious inconsistency in the direct estimates as there was substantial unexplained heterogeneity (I2≥50%, P<0.10). 3We rated down the evidence for serious inconsistency in the network estimates as there was evidence of incoherence (the 95% CIs of the direct

and indirect estimates were significantly different, P<0.05). 4We rated down the evidence for serious indirectness in the direct estimates as only one RCT that was not sufficiently generalizable was available

for assessment. 5We rated down the evidence for serious indirectness in the indirect estimates as there were differences in study characteristics that may modify treatment effect in the direct comparisons (intransitivity). 6We rated down the evidence for serious imprecision in the direct and/or network estimates as the 95% CIs were wide and crossed prespecified

MIDs for benefit and/or harm. MIDs for each outcome were: 0.5kg for body weight, 0.2kg/m2 for BMI, 2% for body fat and 2cm for waist circumference. 7The certainty of the evidence for the network was based on the certainty of the evidence for the indirect estimate as no direct evidence was

available. cPuhan, M.A., et al., The GRADE Working Group (2014). A GRADE Working Group Approach for Rating the Quality of Treatment Effect

Estimates from Network Meta-Analysis. BMJ 349. doi:10.1136/bmj.g5630. dBrignardello-Petersen, R., et al., The GRADE Working Group (2018). Advances in the GRADE Approach to Rate the Certainty in Estimates from a Network Meta-Analysis. Journal of Clinical Epidemiology 93. doi:10.1016/j.jclinepi.2017.10.005.

77



on Glycemic Outcomes

Network analyses of randomized controlled trials investigating the effect of the substitution of NSBs for SSBs on glycemic outcomes. Data was

pooled using network random effects models and expressed as mean differences (MD) and 95% confidence intervals (CIs). MDs for the direct

pair-wise differences were synthesized with the use of generic inverse variance random effects models. MDs for the indirect pair-wise differences were synthesized with the use of generic inverse variance random effects models and 95% CIs using a common comparator. To display the

results for outcomes on the same plot, standardized mean differences (SMDs) and pseudo 95% CIs (proportionally scaled to the 95% CIs of the

MDs) were calculated. SMDs are represented by a red diamond and 95% CIs by the black horizontal line. Filled, black boxes represent which categories downgrades were made among the direct, indirect and network estimates. The starting point for the certainty of the evidence for the

network estimate was based on the certainty of the evidence for the direct or indirect estimate that contributed the most weight to the network

estimate. 2h-PG, two-hour post prandial glucose; FPG, fasting plasma glucose; FPI, fasting plasma insulin; GRADE, Grading of Recommendations,

Assessment, Development, and Evaluation; HbA1c, hemoglobin A1c; HOMA-IR, Homeostatic Model Assessment of Insulin Resistance; MID,

Minimally Important Difference; NE, not estimable; No., number; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages. 1The certainty of evidence for the indirect estimates is based off the lowest rating in the first order loop. 2We rated down the evidence for serious inconsistency in the direct estimates as there was substantial unexplained heterogeneity (I2≥50%,

P<0.10). 3We rated down the evidence for serious inconsistency in the network estimates as there was evidence of incoherence (the 95% CIs of the direct

and indirect estimates were significantly different, P<0.05). 4We rated down the evidence for serious indirectness in the direct estimates as only one RCT that was not sufficiently generalizable was available for assessment. 5We rated down the evidence for serious indirectness in the indirect estimates as there were differences in study characteristics that may modify

treatment effect in the direct comparisons (intransitivity). 6We rated down the evidence for serious imprecision in the direct and/or network estimates as the 95% CIs were wide and crossed prespecified

MIDs for benefit and/or harm. MIDs for each outcome were: 0.3% for HbA1c, 0.5mmol/L for FPG, 10% for 2h-PG, 5pmol for FPI and 1 for

HOMA-IR. cPuhan, M.A., et al., The GRADE Working Group (2014). A GRADE Working Group Approach for Rating the Quality of Treatment Effect

Estimates from Network Meta-Analysis. BMJ 349. doi:10.1136/bmj.g5630. dBrignardello-Petersen, R., et al., The GRADE Working Group (2018). Advances in the GRADE Approach to Rate the Certainty in Estimates from a Network Meta-Analysis. Journal of Clinical Epidemiology 93. doi:10.1016/j.jclinepi.2017.10.005.

78



on Blood Lipid Outcomes

Network analyses of randomized controlled trials investigating the effect of the substitution of NSBs for SSBs on blood lipid outcomes. Data was






estimate. GRADE, Grading of Recommendations, Assessment, Development, and Evaluation; HDL-C, high-density lipoprotein-cholesterol; LDL-C, low-

density lipoprotein-cholesterol; MID, Minimally Important Difference; No., number; Non-HDL-C, non-high-density lipoprotein-cholesterol;

NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages; TC, total cholesterol; TGs, triglycerides. 1The certainty of evidence for the indirect estimates is based off the lowest certainty of evidence in the first order loop. 2We downgraded the evidence for serious inconsistency in the direct estimates as there was substantial unexplained heterogeneity (I2≥50%,

P<0.10). 3We downgraded the evidence for serious inconsistency in the network estimates as there was evidence of incoherence (the 95% CIs of the direct

and indirect estimates were significantly different, P<0.05). 4We downgraded the evidence for serious indirectness in the direct estimates as only one RCT that was not sufficiently generalizable was available for assessment. 5We downgraded the evidence for serious indirectness in the indirect estimates as there were differences in study characteristics that may modify

treatment effect in the direct comparisons (intransitivity). 6We downgraded the evidence for serious imprecision in the direct and/or network estimates as the 95% CIs were wide and crossed prespecified

MIDs for benefit and/or harm. MIDs for each outcome were: 0.1mmol/L. cPuhan, M.A., et al., The GRADE Working Group (2014). A GRADE Working Group Approach for Rating the Quality of Treatment Effect Estimates from Network Meta-Analysis. BMJ 349. doi:10.1136/bmj.g5630. dBrignardello-Petersen, R., et al., The GRADE Working Group (2018). Advances in the GRADE Approach to Rate the Certainty in Estimates

from a Network Meta-Analysis. Journal of Clinical Epidemiology 93. doi:10.1016/j.jclinepi.2017.10.005.

79



on Blood pressure, NAFLD and Uric Acid Outcomes

Network analyses of randomized controlled trials investigating the effect of the substitution of NSBs for SSBs on blood pressure, NAFLD and

uric acid outcomes. Data was pooled using network random effects models and expressed as mean differences (MD) and 95% confidence

intervals (CIs). MDs for the direct pair-wise differences were synthesized with the use of generic inverse variance random effects models. MDs for the indirect pair-wise differences were synthesized with the use of generic inverse variance random effects models and 95% CIs using a

common comparator. To display the results for outcomes on the same plot, standardized mean differences (SMDs) and pseudo 95% CIs

(proportionally scaled to the 95% CIs of the MDs) were calculated. SMDs are represented by a red diamond and 95% CIs by the black horizontal line. Filled, black boxes represent which categories downgrades were made among the direct, indirect and network estimates. The starting point

for the certainty of the evidence for the network estimate was based on the certainty of the evidence for the direct or indirect estimate that

contributed the most weight to the network estimate. ALT, alanine transaminase; AST, aspartate transaminase; DBP, diastolic blood pressure; GRADE, Grading of Recommendations, Assessment,

Development, and Evaluation; IHCL, intra-hepatocellular lipid; MID, Minimally Important Difference; NE, not estimable; No., number; NSBs,

non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages; SBP, systolic blood pressure. 1The certainty of evidence for the indirect estimates is based off the lowest certainty of evidence in the first order loop. 2We downgraded the evidence for serious inconsistency in the direct estimates as there was substantial unexplained heterogeneity (I2≥50%,




MIDs for benefit and/or harm. MIDs for each outcome were: 2mmHg for SBP, 2mmHg for DBP, 10% for IHCL, 10% for ALT, 10% for AST,

and 0.22 mg/dl (0.013mmol/L) for uric acid. cPuhan, M.A., et al., The GRADE Working Group (2014). A GRADE Working Group Approach for Rating the Quality of Treatment Effect

Estimates from Network Meta-Analysis. BMJ 349. doi:10.1136/bmj.g5630. dBrignardello-Petersen, R., et al., The GRADE Working Group (2018). Advances in the GRADE Approach to Rate the Certainty in Estimates

from a Network Meta-Analysis. Journal of Clinical Epidemiology 93. doi:10.1016/j.jclinepi.2017.10.005d Brignardello-Petersen, R., et al.,

The GRADE Working Group (2018). Advances in the GRADE Approach to Rate the Certainty in Estimates from a Network

Meta-Analysis. Journal of Clinical Epidemiology 93. doi:10.1016/j.jclinepi.2017.10.005

80




Network analyses of randomized controlled trials investigating the effect of the substitution of water for SSBs on outcomes of adiposity. Data was pooled using network random effects models and expressed as mean differences (MD) and 95% confidence intervals (CIs). MDs for the

direct pair-wise differences were synthesized with the use of generic inverse variance random effects models. MDs for the indirect pair-wise

differences were synthesized with the use of generic inverse variance random effects models and 95% CIs using a common comparator. To display the results for outcomes on the same plot, standardized mean differences (SMDs) and pseudo 95% CIs (proportionally scaled to the 95%

CIs of the MDs) were calculated. SMDs are represented by a red diamond and 95% CIs by the black horizontal line. Filled, black boxes represent

which categories downgrades were made among the direct, indirect and network estimates. The starting point for the certainty of the evidence for the network estimate was based on the certainty of the evidence for the direct or indirect estimate that contributed the most weight to the network

estimate.

BMI, body mass index; GRADE, Grading of Recommendations, Assessment, Development, and Evaluation; MID, Minimally Important Difference; NE, not estimable; No., number; SSBs, sugar-sweetened beverages; WC, waist circumference. 1The certainty of evidence for the indirect estimates is based off the lowest certainty of evidence in the first order loop. 2We downgraded the evidence for serious inconsistency in the direct estimates as there was substantial unexplained heterogeneity (I2≥50%,

P<0.10). 3We downgraded the evidence for serious inconsistency in the network estimates as there was evidence of incoherence (the 95% CIs of the direct and indirect estimates were significantly different, P<0.05). 4We downgraded the evidence for serious indirectness in the direct estimates as only one RCT that was not sufficiently generalizable was

available for assessment. 5We downgraded the evidence for serious indirectness in the indirect estimates as there were differences in study characteristics that may modify


MIDs for benefit and/or harm. MIDs for each outcome were: 0.5kg for body weight, 0.2kg/m2 for BMI, 2% for body fat and 2cm for waist

circumference. cPuhan, M.A., et al., The GRADE Working Group (2014). A GRADE Working Group Approach for Rating the Quality of Treatment Effect Estimates from Network Meta-Analysis. BMJ 349. doi:10.1136/bmj.g5630. dBrignardello-Petersen, R., et al., The GRADE Working Group (2018). Advances in the GRADE Approach to Rate the Certainty in Estimates


81



on Glycemic Outcomes

Network analyses of randomized controlled trials investigating the effect of the substitution of water for SSBs on glycemic outcomes. Data was






estimate. 2h-PG, two-hour post prandial glucose; FPG, fasting plasma glucose; FPI, fasting plasma insulin; GRADE, Grading of Recommendations,


Minimally Important Difference; NE, not estimable; No., number; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages. 1The certainty of evidence for the indirect estimates is based off the lowest certainty of evidence in the first order loop. 2We downgraded the evidence for serious inconsistency in the direct estimates as there was substantial unexplained heterogeneity (I2≥50%,





HOMA-IR. 7The certainty of the evidence for the network was based on the certainty of the evidence for the indirect estimate as no direct evidence was

available. cPuhan, M.A., et al., The GRADE Working Group (2014). A GRADE Working Group Approach for Rating the Quality of Treatment Effect Estimates from Network Meta-Analysis. BMJ 349. doi:10.1136/bmj.g5630. dBrignardello-Petersen, R., et al., The GRADE Working Group (2018). Advances in the GRADE Approach to Rate the Certainty in Estimates


82




Network analyses of randomized controlled trials investigating the effect of the substitution of water for SSBs on blood lipid outcomes. Data was

pooled using network random effects models and expressed as mean differences (MD) and 95% confidence intervals (CIs). MDs for the direct pair-wise differences were synthesized with the use of generic inverse variance random effects models. MDs for the indirect pair-wise differences

were synthesized with the use of generic inverse variance random effects models and 95% CIs using a common comparator. To display the

results for outcomes on the same plot, standardized mean differences (SMDs) and pseudo 95% CIs (proportionally scaled to the 95% CIs of the MDs) were calculated. SMDs are represented by a red diamond and 95% CIs by the black horizontal line. Filled, black boxes represent which

categories downgrades were made among the direct, indirect and network estimates. The starting point for the certainty of the evidence for the

network estimate was based on the certainty of the evidence for the direct or indirect estimate that contributed the most weight to the network estimate.

GRADE, Grading of Recommendations, Assessment, Development, and Evaluation; HDL-C, high-density lipoprotein-cholesterol; LDL-C, low-

density lipoprotein-cholesterol; MID, Minimally Important Difference; No., number; Non-HDL-C, non-high-density lipoprotein-cholesterol; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages; TC, total cholesterol; TGs, triglycerides. 1The certainty of evidence for the indirect estimates is based off the lowest certainty of evidence in the first order loop. 2We downgraded the evidence for serious inconsistency in the direct estimates as there was substantial unexplained heterogeneity (I2≥50%, P<0.10). 3We downgraded the evidence for serious inconsistency in the network estimates as there was evidence of incoherence (the 95% CIs of the direct

and indirect estimates were significantly different, P<0.05). 4We downgraded the evidence for serious indirectness in the direct estimates as only one RCT that was not sufficiently generalizable was

available for assessment. 5We downgraded the evidence for serious indirectness in the indirect estimates as there were differences in study characteristics that may modify treatment effect in the direct comparisons (intransitivity). 6We downgraded the evidence for serious imprecision in the direct and/or network estimates as the 95% CIs were wide and crossed prespecified

MIDs for benefit and/or harm. MIDs for each outcome were: 0.1mmol/L. cPuhan, M.A., et al., The GRADE Working Group (2014). A GRADE Working Group Approach for Rating the Quality of Treatment Effect



83




Network analyses of randomized controlled trials investigating the effect of the substitution of water for SSBs on blood pressure, NAFLD and

uric acid outcomes. Data was pooled using network random effects models and expressed as mean differences (MD) and 95% confidence intervals (CIs). MDs for the direct pair-wise differences were synthesized with the use of generic inverse variance random effects models. MDs

for the indirect pair-wise differences were synthesized with the use of generic inverse variance random effects models and 95% CIs using a

common comparator. To display the results for outcomes on the same plot, standardized mean differences (SMDs) and pseudo 95% CIs (proportionally scaled to the 95% CIs of the MDs) were calculated. SMDs are represented by a red diamond and 95% CIs by the black horizontal

line. Filled, black boxes represent which categories downgrades were made among the direct, indirect and network estimates. The starting point

for the certainty of the evidence for the network estimate was based on the certainty of the evidence for the direct or indirect estimate that contributed the most weight to the network estimate.

ALT, alanine transaminase; AST, aspartate transaminase; DBP, diastolic blood pressure; GRADE, Grading of Recommendations, Assessment,

Development, and Evaluation; IHCL, intra-hepatocellular lipid; MID, Minimally Important Difference; NE, not estimable; No., number; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages; SBP, systolic blood pressure. 1The certainty of evidence for the indirect estimates is based off the lowest certainty of evidence in the first order loop. 2We downgraded the evidence for serious inconsistency in the direct estimates as there was substantial unexplained heterogeneity (I2≥50%, P<0.10). 3We downgraded the evidence for serious inconsistency in the network estimates as there was evidence of incoherence (the 95% CIs of the direct



MIDs for benefit and/or harm. MIDs for each outcome were: 2mmHg for SBP, 2mmHg for DBP, 10% for IHCL, 10% for ALT, 10% for AST, and 0.22 mg/dl (0.013mmol/L) for uric acid. 7The certainty of the evidence for the network was based on the certainty of the evidence for the indirect estimate as no direct evidence was

available. c Puhan, M.A., et al., The GRADE Working Group (2014). A GRADE Working Group Approach for Rating the Quality of Treatment Effect

Estimates from Network Meta-Analysis. BMJ 349. doi:10.1136/bmj.g5630. d Brignardello-Petersen, R., et al., The GRADE Working Group (2018). Advances in the GRADE Approach to Rate the Certainty in Estimates from a Network Meta-Analysis. Journal of Clinical Epidemiology 93. doi:10.1016/j.jclinepi.2017.10.005.

84




Network analyses of randomized controlled trials investigating the effect of the substitution of NSBs for water on outcomes of adiposity. Data was pooled using network random effects models and expressed as mean differences (MD) and 95% confidence intervals (CIs). MDs for the

direct pair-wise differences were synthesized with the use of generic inverse variance random effects models. MDs for the indirect pair-wise

differences were synthesized with the use of generic inverse variance random effects models and 95% CIs using a common comparator. To display the results for outcomes on the same plot, standardized mean differences (SMDs) and pseudo 95% CIs (proportionally scaled to the 95%

CIs of the MDs) were calculated. SMDs are represented by a red diamond and 95% CIs by the black horizontal line. Filled, black boxes represent

which categories downgrades were made among the direct, indirect and network estimates. The starting point for the certainty of the evidence for the network estimate was based on the certainty of the evidence for the direct or indirect estimate that contributed the most weight to the network

estimate.

BMI, body mass index; GRADE, Grading of Recommendations, Assessment, Development, and Evaluation; MID, Minimally Important Difference; NE, not estimable; No., number; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages; WC waist

circumference. 1The certainty of evidence for the indirect estimates is based off the lowest certainty of evidence in the first order loop. 2We downgraded the evidence for serious inconsistency in the direct estimates as there was substantial unexplained heterogeneity (I2≥50%,

P<0.10). 3We downgraded the evidence for serious inconsistency in the network estimates as there was evidence of incoherence (the 95% CIs of the direct and indirect estimates were significantly different, P<0.05). 4We downgraded the evidence for serious indirectness in the direct estimates as only one RCT that was not sufficiently generalizable was


treatment effect in the direct comparisons (intransitivity). 6We downgraded the evidence for serious imprecision in the direct and/or network estimates as the 95% CIs were wide and crossed prespecified MIDs for benefit and/or harm. MIDs for each outcome were: 0.5kg for body weight, 0.2kg/m2 for BMI, 2% for body fat and 2cm for waist

circumference. cPuhan, M.A., et al., The GRADE Working Group (2014). A GRADE Working Group Approach for Rating the Quality of Treatment Effect Estimates from Network Meta-Analysis. BMJ 349. doi:10.1136/bmj.g5630. dBrignardello-Petersen, R., et al., The GRADE Working Group (2018). Advances in the GRADE Approach to Rate the Certainty in Estimates


85



Network analyses of randomized controlled trials investigating the effect of the substitution of NSBs for water on glycemic outcomes. Data was


pair-wise differences were synthesized with the use of generic inverse variance random effects models. MDs for the indirect pair-wise differences



MDs) were calculated. SMDs are represented by a red diamond and 95% CIs by the black horizontal line. Filled, black boxes represent which



estimate.

2h-PG, two-hour post prandial glucose; FPG, fasting plasma glucose; FPI, fasting plasma insulin; GRADE, Grading of Recommendations,


Minimally Important Difference; NE, not estimable; No., number; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages. 1The certainty of evidence for the indirect estimates is based off the lowest certainty of evidence in the first order loop. 2We downgraded the evidence for serious inconsistency in the direct estimates as there was substantial unexplained heterogeneity (I2≥50%,






HOMA-IR. cPuhan, M.A., et al., The GRADE Working Group (2014). A GRADE Working Group Approach for Rating the Quality of Treatment Effect



86




Network analyses of randomized controlled trials investigating the effect of the substitution of NSBs for water on blood lipid outcomes. Data was

pooled using network random effects models and expressed as mean differences (MD) and 95% confidence intervals (CIs). MDs for the direct pair-wise differences were synthesized with the use of generic inverse variance random effects models. MDs for the indirect pair-wise differences


results for outcomes on the same plot, standardized mean differences (SMDs) and pseudo 95% CIs (proportionally scaled to the 95% CIs of the MDs) were calculated. SMDs are represented by a red diamond and 95% CIs by the black horizontal line. Filled, black boxes represent which


network estimate was based on the certainty of the evidence for the direct or indirect estimate that contributed the most weight to the network estimate.

GRADE, Grading of Recommendations, Assessment, Development, and Evaluation; HDL-C, high-density lipoprotein-cholesterol; LDL-C, low-

density lipoprotein-cholesterol; MID, Minimally Important Difference; No., number; Non-HDL-C, non-high-density lipoprotein-cholesterol; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages; TC, total cholesterol; TGs, triglycerides. 1The certainty of evidence for the indirect estimates is based off the lowest certainty of evidence in the first order loop. 2We downgraded the evidence for serious inconsistency in the direct estimates as there was substantial unexplained heterogeneity (I2≥50%, P<0.10). 3We downgraded the evidence for serious inconsistency in the network estimates as there was evidence of incoherence (the 95% CIs of the direct



MIDs for benefit and/or harm. MIDs for each outcome were: 0.1mmol/L. cPuhan, M.A., et al., The GRADE Working Group (2014). A GRADE Working Group Approach for Rating the Quality of Treatment Effect



87




Network analyses of randomized controlled trials investigating the effect of the substitution of NSBs for water on blood pressure, NAFLD and

uric acid outcomes. Data was pooled using network random effects models and expressed as mean differences (MD) and 95% confidence

intervals (CIs). MDs for the direct pair-wise differences were synthesized with the use of generic inverse variance random effects models. MDs

for the indirect pair-wise differences were synthesized with the use of generic inverse variance random effects models and 95% CIs using a

common comparator. To display the results for outcomes on the same plot, standardized mean differences (SMDs) and pseudo 95% CIs

(proportionally scaled to the 95% CIs of the MDs) were calculated. SMDs are represented by a red diamond and 95% CIs by the black horizontal

line. Filled, black boxes represent which categories downgrades were made among the direct, indirect and network estimates. The starting point

for the certainty of the evidence for the network estimate was based on the certainty of the evidence for the direct or indirect estimate that

contributed the most weight to the network estimate.

ALT, alanine transaminase; AST, aspartate transaminase; DBP, diastolic blood pressure; GRADE, Grading of Recommendations, Assessment,

Development, and Evaluation; IHCL, intra-hepatocellular lipid; MID, Minimally Important Difference; NE, not estimable; No., number; NSBs,

non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages; SBP, systolic blood pressure. 1The certainty of evidence for the indirect estimates is based off the lowest certainty of evidence in the first order loop. 2We downgraded the evidence for serious inconsistency in the direct estimates as there was substantial unexplained heterogeneity (I2≥50%,





MIDs for benefit and/or harm. MIDs for each outcome were: 2mmHg for SBP, 2mmHg for DBP, 10% for IHCL, 10% for ALT, 10% for AST,

and 0.22 mg/dl (0.013mmol/L) for uric acid. cPuhan, M.A., et al., The GRADE Working Group (2014). A GRADE Working Group Approach for Rating the Quality of Treatment Effect



88

23Appendix Figure 4.19: Network Diagram for Randomized Controlled Trials Investigating the

Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on Body

Weight

Blue nodes represent the study size for each beverage under investigation. The thickness of the black lines

represents the number of studies directly comparing one beverage to another. NSBs, non-nutritive sweetened

beverages; SSBs, sugar-sweetened beverages.


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on BMI


represents the number of studies directly comparing one beverage to another. BMI; body mass index, NSBs, non-

nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

89


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on Body Fat

%





Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on Waist

Circumference




90


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on HbA1c

Blue nodes represent the study size for each beverage under investigation. The thickness of the

black lines represents the number of studies directly comparing one beverage to another. HbA1c,

hemoglobin A1c; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on FPG


represents the number of studies directly comparing one beverage to another. FPG, fasting plasma glucose; NSBs,

non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

91


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on 2h-PG


represents the number of studies directly comparing one beverage to another. 2h-PG, two-hour post prandial

glucose; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on FPI


represents the number of studies directly comparing one beverage to another. FPI, fasting plasma insulin; NSBs,


92


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on HOMA-IR


represents the number of studies directly comparing one beverage to another. HOMA-IR; homeostatic model

assessment of insulin resistance; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on LDL-C


represents the number of studies directly comparing one beverage to another. LDL-C, low-density lipoprotein

cholesterol; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

93


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on Non-HDL-

C


represents the number of studies directly comparing one beverage to another. Non-HDL-C, non-high-density

lipoprotein cholesterol, NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on TGs



beverages; SSBs, sugar-sweetened beverages; TGs, triglycerides.

94


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on HDL-C


represents the number of studies directly comparing one beverage to another. HDL-C, high-density lipoprotein

cholesterol, NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages.

36Appendix l Figure 4.32: Network Diagram for Randomized Controlled Trials Investigating the

Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on TC



beverages; SSBs, sugar-sweetened beverages; TC, total cholesterol.

95


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on SBP



beverages; SBP, systolic blood pressure; SSBs, sugar-sweetened beverages.


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on DBP


represents the number of studies directly comparing one beverage to another. DBP, diastolic blood pressure; NSBs,


96

39Appendix Figure 4.35 Network Diagram for Randomized Controlled Trials Investigating the

Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on IHCL


represents the number of studies directly comparing one beverage to another. IHCL, intra-hepatocellular lipid;



Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on ALT


represents the number of studies directly comparing one beverage to another. ALT, alanine transaminase; NSBs,


97


Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on AST


represents the number of studies directly comparing one beverage to another. AST, aspartate transaminase; NSBs,



Effect of the Substitution of NSBs for SSBs, Water for SSBs, and NSBs for Water on uric acid




98

Chapter 5: Rationale, Design and Baseline Characteristics Assessing the

Effect of Substituting NSBs versus Water for SSBs on Gut Microbiome,

Glucose Tolerance, and Cardiometabolic Risk Factors: Strategies To

OPpose SUGARS with Non-nutritive sweeteners Or Water trial (STOP

Sugars NOW)

Néma McGlynn1,2,3, Sabrina Ayoub-Charette1,2,3, Tauseef A Khan1,2,3, Sonia Blanco-Mejia1,2,3, Laura Chiavaroli1,2,3,

Meaghan Kavanagh1,2,3, Danielle Lee1,2,3, Elena M Comelli3,4, Cyril WC Kendall1,2,3,5, Lawrence A Leiter1,2,3,6,7,8,

John L Sievenpiper1,2,3,6,7,8

1Toronto 3D Knowledge Synthesis and Clinical Trials Unit, Toronto, ON, Canada; 2Clinical Nutrition and Risk Factor Modification Centre, St. Michael's Hospital, Toronto, ON, Canada; 3Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, ON, Canada 4Joannah and Brian Lawson Centre for Child Nutrition, Toronto, ON, Canada. 5College of Pharmacy and Nutrition, University of Saskatchewan, Saskatoon, SK, Canada; 6Division of Endocrinology and Metabolism, Department of Medicine, St. Michael’s Hospital, Toronto, ON,

Canada; 7Department of Medicine, Faculty of Medicine, University of Toronto, Toronto, ON, Canada; 8Li Ka Shing Knowledge Institute, St. Michael’s Hospital, Toronto, ON, Canada;

Corresponding Author: John L Sievenpiper MD, PhD, FRCPC, St. Michael's Hospital, #6138-61 Queen Street East,

Toronto, ON, M5C 2T2, CANADA, Tel: 416 867 7475, Fax: 416 867 7495, email: [email protected]

mailto:[email protected]

99

5.1 Abstract

Background: Concerns exist that non-nutritive sweetened beverages (NSBs) do not have

established benefits and adversely affect glucose tolerance through compositional changes of the

gut microbiome. Whether the intended benefits of NSBs improve glucose control similar to

water in their substitution for sugar-sweetened beverages (SSBs) is unclear. The STOP Sugars

NOW study is a randomized crossover trial on the effect of 3 prespecified substitutions (NSBs

for SSBs, water for SSBs and NSBs for water) on gut microbiome and glucose tolerance. This

report provides the rationale, design and baseline characteristics of the trial.

Methods: A randomized crossover trial with three 4-week intervention phases (SSBs, NSBs and

water) in 81 overweight and obese adult participants, who consume at least one 355mL can of

SSBs daily. Each intervention phase is separated by a ≥4-week washout phase where participants

revert back to their regular SSB intake. The protocol includes 6 study visits. Two primary

outcomes, gut microbiome beta-diversity and glucose tolerance, are analyzed through 16S rRNA

sequencing of fecal sample collections and 2-hour 75g oral glucose tolerance tests, respectively.

Adherence to study beverages is assessed through biomarker analysis of 24-hour urine, collected

at each study visit.

Results: Participants were mainly recruited through Toronto Transit Commission subway

advertisements and a third-party participant recruitment service (Trialfacts). At baseline,

participants had a mean age of 41.8 ± 12.9 years, weight of 94.2 ± 18.9kg, BMI of 33.8 ±

6.7kg/m2, and waist circumference of 108.7 ± 13.4cm. Mean baseline levels of blood pressure

were normal. STOP Sugars NOW participants consumed an average of two 355mL cans of SSBs

daily, 43.2% consumed Coke, and 25.9% consumed Canada Dry Ginger Ale. Projected NSB

intake indicates 95% of participants will be consuming a blend of aspartame and acesulfame

potassium. Sixty-one percent of participants reported no consumption of any foods or beverages

with non-nutritive sweeteners (NNSs) over the last 6 months at baseline, whereas 24.7% of those

who consumed NNSs did so from NSBs.

Conclusion: The STOP Sugars NOW trial will be the first randomized controlled trial to

investigate the effect of 3 prespecified substitutions (NSBs for SSBs, water for SSBs and NSBs

for water) on changes in the gut microbiota and glucose tolerance.

Trial registration number: ClinicalTrials.gov identifier, NCT03543644

100

5.2 Introduction

Health authorities discourage the consumption of sugar-sweetened beverages (SSBs) [11, 73,

186, 187] as their excess intakes have been associated with weight gain, type 2 diabetes mellitus

(T2DM), and their downstream complications of hypertension and coronary heart disease (CHD)

[6-9, 188].

Non-nutritive sweetened beverages (NSBs) provide a viable alternative for SSBs, yet concerns

exist that they do not have established benefits, with mixed recommendations for NSBs in

dietary and clinical practice guidelines [11, 12, 15, 16, 18], the majority recommending that

water replace SSBs. Whether the intended benefits of NSBs improve glucose control similar to

water in their substitution for SSBs is unclear.

Systematic reviews and meta-analyses of prospective cohort studies, including a WHO-

commissioned review [21], indicate that NSBs are associated with an increased risk of the

conditions they are intended to prevent including weight gain, type 2 diabetes mellitus (T2DM)

and coronary heart disease (CHD) [7, 19], even though reverse causality and confounding by

indication are the most likely explanations [134, 143].

Syntheses of randomized controlled trials (RCTs), have failed to fully account for the calories

available to be displaced by NSBs with caloric (e.g. SSBs) and noncaloric (e.g. water, placebo,

weight loss diets) comparators pooled together, or noncaloric comparators used as the sole

comparator leading to an underestimation of the effect of NSBs in their intended substitution for

SSBs [19-21].

Although biological mechanisms involving impaired sensory and endocrine signaling mediated

by the sweet taste receptor [79, 80] have been offered in support of these observations, there is

particular concern that NSBs may induce and promote glucose intolerance through changes to

the microbiome [130, 131]. Much of this concern derives from a single intervention study that

showed a decrease in glucose tolerance after 6 days of saccharin in an NSB format with

subsequent transfer of the phenotype to germ free mice by fecal transplant [81]. This study,

despite its concerning methodological weaknesses, including a small sample size of 7 healthy

participants, daily continuous 75g Oral Glucose Tolerance Tests (OGTT) and no control group,

101

resulted in an overall negative view of non-nutritive sweeteners (NNSs) in the media [182, 183,

189-194].

These prevailing uncertainties signal a need to address the ongoing concerns related to NSBs.

Health Canada, in particular, has indicated that studies of sugar reduction strategies that use

NNSs and target the microbiome are an important research priority [195]. We have therefore,

designed and implemented a Canadian Institutes for Health Research (CIHR)-funded RCT to

assess the effect of a “real world” strategy of 3 prespecified substitutions: NSBs for SSBs

(intended substitution with caloric displacement), water for SSBs (“standard of care” substitution

with caloric displacement) and NSBs for water (matched substitution without caloric

displacement) on gut microbiome and glucose tolerance in overweight/obese participants: The

STOP Sugars NOW trial.

5.3 Objective and Hypotheses

There are 3 primary objectives of this study.

1. To assess the effect of our 3 prespecified substitutions (NSBs for SSBs, water for SSBs

and NSBs for water) on the first primary outcome of diversity of gut microbiome over 4-

weeks in overweight/obese participants who are regular SSB drinkers.

2. To assess the effect of our 3 prespecified substitutions (NSBs for SSBs, water for SSBs

and NSBs for water) on the second primary outcome of glucose tolerance over 4-weeks

in overweight/obese participants who are regular SSB drinkers.

3. To assess the effect our 3 prespecified substitutions (NSBs for SSBs, water for SSBs and

NSBs for water) on the secondary outcomes of body weight, body mass index (BMI),

waist circumference and glucose and insulin regulation over 4-weeks in overweight/obese

participants who are regular SSB drinkers.

The hypothesis for each objective is listed below.

1. Substituting NSBs for SSBs and water for SSBs will improve the diversity of gut

microbiome after 4-weeks in overweight/obese participants who are regular SSB

drinkers, with no difference between the substitution of NSBs for water.

102

2. Substituting NSBs for SSBs and water for SSBs will improve glucose tolerance after 4-

weeks in overweight/obese participants who are regular SSB drinkers, with no difference

between the substitution of NSBs for water.

3. Substituting NSBs for SSBs and water for SSBs will decrease measures of adiposity and

improve glucose and insulin regulation after 4-weeks in overweight/obese participants

who are regular SSB drinkers, with no difference between the substitution of NSBs for

water.

5.4 Methods

5.4.1 Study Design

The STOP Sugars NOW trial is a single-center, open label, randomized controlled crossover trial

among 81 overweight or obese participants. This study was approved by The Unity Health

Toronto Research Ethics Board in Jan 2018 and is registered at www.clinicaltrials.gov

(NCT03543644).

In order to compare the effect of our 3 prespecified substitutions on gut microbiome diversity

and glucose tolerance, participants consume SSBs, NSBs or water over a period of 4-weeks each.

Due to the high inter-individual heterogeneity in the gut microbiome, a crossover design was

selected [196]. A schematic representation of the study design can be seen in Figure 5.1.

Participants start the study as their own control receiving the 3 interventions in random order,

with each separated by a ≥4-week washout phase.

103

5.4.2 Blinding

Blinding of the participants and investigators is not possible due to the packaging, taste and look

of the interventions. However, outcome assessors (laboratory, microbiome analysis) and the

statistician will be blinded to the identity of the treatments.

5.4.3 Participants

We recruited 81 participants consuming ≥1 can (355ml) serving of SSBs per day (for which

there is an NSB equivalent that contains the NNSs acesulfame potassium (Ace-K) or sucralose),

who were overweight or obese based on ethnic-specific cut-offs (BMI ≥23kg/m2 for Asian

individuals and ≥ 25kg/m2 other individuals [37]), with a high waist circumference (≥94cm in

men, ≥80cm in women of Europid, Sub-Saharan African, Eastern Mediterranean, and Middle

Eastern origin; ≥90cm in men and ≥ 80cm in women of South Asian, Chinese, Japanese, and

South and Central American origin [35]) between the ages of 18-75 years, non-smokers, not

taking medications that have a clinically relevant effect on the primary outcomes, no intentions

of making major alterations to their diet or physical activity regime during the study duration,

had a primary care physician, and were proficient in English were eligible to participate in the

trial. The full set of exclusion criteria can be found on Table 5.1. Participants had to agree to

consume only one type of SSB and brand-matched NSB for each month of intervention. If

participants were consuming NSBs in addition to SSBs they were asked to discontinue

consuming them for the run-in and duration of the study. All participants are expected to

participate in the study for approximately six months (one run-in phase of ≥2-weeks, three

intervention phases of 4-weeks each, and two washout phases of ≥4-weeks each) and be able to

attend all study visits.

104

5.4.4 Sample Size (Power) Calculation

The study is being performed in 81 participants and powered to show a difference between the

NSB, water and SSB arms in 60 participants in the two primary outcomes. Assuming a drop-out

rate of 20%, we recruited 75 participants. Our final recruitment strategy resulted in a larger

number of participants who completed the run-in phase. Consequently, we increased our

recruitment to replace drop-outs, resulting in the recruitment 81 participants. Changes in beta

diversity of gut microbiome and changes in glucose tolerance are the two primary outcomes of

the trial. It is hypothesized that substituting NSBs for SSBs and water for SSBs will increase the

diversity of the gut microbiome and improve glucose tolerance after 4-weeks in regular SSB

consumers, with no difference between the substitution of NSBs for water. We are powered to

detect a difference between all three interventions (based on a power calculation difference

between the NSB and water arm), and this difference is expected to be smaller than the NSB for

SSB and water for SSB comparisons.

The first primary outcome of gut microbiome beta diversity is being analyzed through 16S

ribosomal rRNA gene sequencing using pairwise weighted UniFrac distances [197]. UniFrac is a

distance metric based upon the unique fraction of branch length in a phylogenetic tree built from

two sets of taxa. Comparison of microbiome samples will be performed via weighted UniFrac,

which considers the relative abundance of taxa. We simulated the within-group distance as 0.2,

and the standard deviation (SD) of within-group distances as 0.07. Previous studies have shown

that diversity of the gut microbiome taxa can change significantly with small dietary alterations

over a short period (5 to 7 days) in small groups of people (10 to 25) [196, 198]. Therefore, a

weighted UniFrac distance of 0.04 was set based on the effect of 0.5 observed in the study by

Suez et al. [81]. Due to the crossover design of this trial with a within-person correlation of 0.7,

it was calculated to show a difference between the water and NSB arms in both primary

outcomes with 98% power, 75 participants were required, assuming a drop-out rate of 20%.

The second primary outcome is change in glucose tolerance, as measured by incremental Area

Under the Curve (iAUC) from a 2-hour 75g OGTT. Assuming absolute numbers for mean and

SD from a recent unpublished randomized trial [199] and a within-person correlation of 0.7, 60

participants provide 89% power to detect a 20% change in mean iAUC between the three arms,

if the direction of change is similar to Suez et al. [81]. The 20% difference for glucose iAUC was

105

based on the minimally important difference proposed by Health Canada for postprandial

glycaemic response health claims [200] (Table 5.2).

To account for the adjustment required for multiple testing of two primary outcomes, this power

calculation uses the Benjamini-Hochberg procedure as suggested by the Food and Drug

Administration’s draft document “Multiple Endpoints in Clinical Trials Guidance for Industry”

[201]. This procedure is a step-down method that controls for false discovery rate, while

maintaining high power [202, 203]. To show a difference between SSBs, NSBs and water for all

primary and secondary outcomes, a truncated Benjamini-Hochberg method with parallel

gatekeeping is being implemented, whereby an unused portion of the alpha (α) is passed onto the

secondary family of outcomes if either of the primary outcomes is significant. An α of 0.0375 for

both primary outcomes were calculated based on a p-value of 0.05. For all secondary outcomes,

an α of 0.0125 was used as it’s the lowest possible starting α for secondary outcomes as per the

truncated Benjamini-Hochberg procedure.

5.4.5 Recruitment, Consent and Screening

Participants were recruited through two main advertising campaigns: The Toronto Transit

Commission (TTC) subway advertisements and Trialfacts, a third-party participant recruitment

service for clinical trials. Some participants completed an online Survey Monkey to assess

general eligibility. Other forms of recruitment included online advertising through Craigslist and

Kijiji and postcards on the University of Toronto, St George, campus and St Michael’s

Hospital’s (SMH) elevators in Toronto, Ontario, Canada. Interested individuals completed a

telephone-screen using a questionnaire and were briefed on the details of the study, treatments,

and procedures.

Eligible and interested individuals were asked to come to the Clinical Nutrition and Risk Factor

Modification Centre (CNRFMC) at SMH for an in-person consent review. Once consent was

obtained, individuals received a screening identification (ID) number and an in-person screening

was scheduled to obtain anthropometric measures such as height, weight and waist

circumference, completing a questionnaire regarding SSB intake, and blood pressure

measurements to confirm study qualification. Blood pressure was taken by Automated Office

Blood Pressure (AOBP) measurements on both arms. The arm with the highest mean systolic

measurements was used for all study visits. If the average of 3 AOBP measurements displayed a

106

systolic blood pressure ≥160mmHg or diastolic blood pressure ≥100mmHg, the individual was

excluded for uncontrolled hypertension [204].

Significant results were provided to participant’s family physician for further follow-up. Those

who met all eligibility criteria were enrolled in the run-in phase of the study and assigned a study

ID number.

5.4.6 Run-In Phase

Prior to randomization, participants underwent a ≥2-week run-in phase to determine their usual

SSB intake and their beverage preference options for SSBs, NSBs and water (flat or carbonated).

Before starting the run-in phase, participants were scheduled for their first study visit. They were

instructed on how to collect a 24-hour urine sample and fecal sample and provided kits to do so

for their first study visit.

5.4.7 Randomization

Randomization, with no stratification, was done using the Research Electronic Data Capture

(REDCap) program by the Applied Health Research Centre (AHRC) after successful completion

of the run-in phase and first study visit. Participants were randomly allocated to six possible

SSB, NSB and water groups using blocked (Latin squares) randomization with a similar number

of participants allocated to each treatment sequence (Table 5.3). The randomization schedule

was created by AHRC through REDCap. Participants were randomized and given their study

beverages once all measures from the first study visit were collected. Allocating participants

through REDCap kept study staff unaware of each participant’s group assignment until revealed

to them at the end of their first visit.

5.4.8 Intervention

There are three interventions: SSBs (355mL, 140kcal, 39g sugars per can); NSBs (355mL, 0kcal,

0g sugars per can); and water (355mL, 0kcal, 0g sugars per can or bottle of still or carbonated

water). The complete list of all available study beverages is listed in Table 5.4. SSBs were

defined as any SSB (sodas and soft drinks) that contains at least 50kcal per 8-oz serving. For the

purpose of the study, SSBs do NOT include fruit drinks, sports/energy drinks, sweetened iced

tea, coffee, or homemade SSBs such as frescas or 100% fruit juice.

107

For each intervention phase, participants are receiving relevant study beverages based on their

group assignment, with instructions to consume only one type of SSB, its brand-matched NSB

and carbonated or flat water for 4-weeks each in random order, while freely consuming their

usual background diet. To allow for “real-world” substitutions using products available on the

market, the calories of the intervention groups are not matched and the number of cans each

participant consumes each day are based off their ≥2-week run-in beverage logs.

5.4.9 Study Visits

All study visits are being conducted at the CFI-funded Toronto 3D Clinical Research Centre and

the CNRFMC at SMH. Participants are completing the three interventions in random order, each

separated by a ≥4-week washout. Motivational phone calls are being made every 2-weeks

between visits to provide participants with reminders and ensure fidelity to the study protocol.

The complete participant visit schedule is outlined in Table 5.5.

After successful completion of the run-in phase, participants came to the study site for their first

visit where their ≥2-week run-in beverage logs were checked for completion and any changes

required were communicated. Participants were notified of their randomization assignments after

completing all baseline tests.

Prior to each study visit, participants are given standard stool and urine collection kits and

instructions on how to complete a weighted three-day diet record (3DDR) which is being

completed and collected for visits 1, 2, 4 and 6 in order to assess adherence to background diet.

Participants are being instructed to collect their fecal sample and 24-hour urine as close to their

visit dates as possible. The fecal kits are required to assess diversity of gut microbiome and the

urine kits for urinary biomarker analysis of adherence. Participants are being informed to collect

their stool samples in the specimen containers provided. If collected prior to a visit date, they are

to store the fecal sample in their personal freezer using the containers and bags provided. All

fecal samples are to arrive at the centre in the bags provided, surrounded with dry ice packs. If

24-hour urine collections are taken ≥1 day before the participants’ study visits, the sample is

being kept in the participant’s fridge until their study visit date. At each visit, study staff retrieve

the stool sample, the 24-hour urine collection, and beverage logs from the participants. Once

samples are received from participants, they are labeled with the participant’s visit number, date

108

and given a fecal randomization code in order to keep study staff blinded for this outcome.

Samples are immediately placed in the lab’s freezer at –20°C until all samples are provided for

that visit. They are then transported with dry ice to the University of Toronto and frozen at –

80°C.

For all six study visits, participants are coming to the study site in a 10 to 12-hour fasted state.

Prior to their visits, participants are being reminded to maintain the same dietary and exercise

patterns the evening before each test and to consume at least 150g of carbohydrate for three non-

consecutive days before their scheduled visit date. Participants are being provided with concrete

examples of what 150g of carbohydrate looks like to ensure instructions are followed. At each

visit, weight, waist circumference and blood pressure measurements are being taken;

questionnaires assessing physical activity levels, NNS intake in foods and beverages and

medication and supplement changes are being administered and a fasting blood sample for

primary and secondary outcomes (including glucose, insulin and biomarkers of adherence) is

taken by an intravenous (IV) nurse. Administration of a 2-hour 75g OGTT is being performed at

each study visit, following standard protocol, to assess glucose tolerance. Between OGTT blood

draws, participants’ weighted 3DDRs and 24-hour urine and fecal sample forms are verified.

After the OGTT participants are given breakfast.

Once all measures are collected, the participant is provided the intervention. Beverage logs are

provided for each of the intervention and washout phases with reminders to log beverage intake

daily. For the washout phases, participants are purchasing their own SSBs. Subjects take home

one week’s worth of their beverages at the first visit of each phase. The rest of their study drinks

for that phase are picked-up by the participant or shipped to their home addresses by Starhawk, a

third-party delivery company. Consent for the shipment or pick-up of study drinks was obtained

in the consent review.

5.4.10 Washout phase

To control for any carry-over effects of one beverage type over another, each of the three

intervention phases is separated by a ≥4-week wash out phase where participants revert back to

their regular SSB intake. Participants are given beverage logs to complete over this ≥4-week

phase. No beverages are provided by the study site during the washout phase. Participants are

109

contacted via phone and email by study staff near the middle and end of each phase to remind

them of their group assignment and the next study visit.

5.4.11 Outcomes

There are two primary outcomes: change in gut microbiome beta diversity as measured by 16S

rRNA gene sequencing, and plasma glucose iAUC measured by a 2-hour 75g OGTT. Secondary

outcomes are changes in waist circumference, body weight, BMI, fasting plasma glucose (FPG),

2h plasma glucose (2h-PG), and the Matsuda whole body insulin sensitivity index (Matsuda

ISIOGTT) [205]. Exploratory outcomes consisting of a range of cardiometabolic/metabolomic

outcomes is also being considered. All outcome assessments are being performed by qualified

study staff and follow standard operating procedures.

5.4.12 Adherence Assessment

Adherence to study drinks is being assessed through completed beverage logs, returned unused

beverage containers, and objective biomarkers of SSBs (increased 13C/12C ratios in serum fatty

acids, increased urinary fructose), NSBs (urinary Ace-K, sucralose, decreased 13C/12C ratios in

serum fatty acids, decreased urinary fructose) and water (decreased 13C/12C ratios in serum

fatty acids, decreased urinary fructose, decreased urinary Ace-K and sucralose) intake.

(see biomarkers of adherence).

To determine adherence to background diet and evaluate changes in diet quality, participants are

being asked to complete a weighted 3DDR over two non-consecutive weekdays and one

weekend day during the run-in phase and last week of each intervention phases. To increase

accuracy of reporting, participants are provided with a food scale during the in-person screen.

Detailed instructions are given, using food models, on how to complete the record by study

dietitians who also review each 3DDR record with participants at their study visits. The 3DDRs

is being analyzed by Nutritics (Dublin, Ireland), a nutrition software tool.

5.4.13 Analytical Techniques

5.4.13.1 Anthropometric Analyses

All anthropometric measurements are being taken while the participant is wearing light clothes

and no shoes. Body weight is being assessed by electronic beam scale (Rice Lake® Weighing

System, Israel, model 240-10). A wall-mounted stadiometer (Perspective Enterprises, Portage,

110

MI, USA) is used to measure height. Waist circumference is being assessed using the Heart and

Stroke Foundation methodology [206] where participants are standing upright with feet

shoulder-width apart. The iliac crest is being determined using the sides of hands and index

fingers. As locating the iliac crest is often challenging among overweight and obese individuals,

to reduce measuring errors, a measuring tape (AMG Medical Inc, vinyl tape measure, 150cm,

AZGAMG116870.) is being used to determine length from the floor up to the top of the iliac

crest for each participant, during the in-person screen. This value is recorded in the participants’

charts and used at each study visit to ensure consistency in waist measurements. Measurements

from the floor to the top of the iliac crest are being done on both the right and left sides of the

body. Waist circumference is measured around the abdomen at the point of the iliac crest.

Participants are instructed to relax by taking in two normal breaths. After the second breath the

tape is to be tightened around the waist, the measurement collected and then recorded in the

participant’s chart. To obtain blood pressure and resting heart rate, participants are first seated in

a quiet, temperature-controlled room for at least 10-15 minutes. Blood pressure is then measured

oscillometrically by AOBP using the OMRON Intellisense HEM-907 according to JNC VII

criteria [207], where an average of three measurements, each separated by one minute, is taken.

5.4.13.2 Biochemical Analyses

The microbiome is being sequenced by next-generation sequencing (Illumina MiSeq) with

primers targeting the appropriate variable V3-V4 region of 16S rRNA [208] for pair-end

sequencing [209]. Sequencing data is being analyzed to assign Operational Taxonomic Units

(OTUs) to determine their relative abundance. Data for selected taxa is being confirmed by

qPCR [208]. Alpha- and beta-diversity indexes is being calculated. The metagenome is being

inferred from compositional data in silico [210]. Microbiome analysis is being performed at the

Department of Nutritional Sciences, University of Toronto.

Plasma samples for glucose and insulin is being separated by centrifuge in the wet laboratory at

the CNRFMC at SMH, within 4 hours. After each study visit, collected blood samples are

transported to the University of Toronto and immediately frozen at –80°C. Analyses of the

plasma glucose is being done at the Banting and Best Diabetes Centre (BBDC) using the glucose

oxidase method [211] at the University of Toronto. Plasma insulin analysis is being done using

enzyme-linked immunosorbent assay (ELISA), a solid phase two-site enzyme immunoassay

111

method using Mercodia AB Insulin ELISA (Uppsala, Sweden) kits [212]. The Thermo

ScientificTM MultiskcanTM FC Microplate Photometer plate with SkanItTM software (Ratastie,

Finland) is being used to process insulin data. Insulin analysis is taking place at the Guelph

Research and Development Centre, Ontario, Canada. Plasma glucose and insulin curves are

being plotted as the incremental change over time and iAUC is being calculated geometrically

for each participant using the trapezoid method, ignoring areas below the fasting value [213].

The Matsuda ISI (Matsuda ISIOGTT) is being calculated using the 75g OGTT derived plasma

glucose (PG) and insulin (PI) values, according to the formula by Matsuda et al. [205]: 10,000

divided by the square root of ([FPG x FPI] x [mean PG x PI insulin during OGTT]), where PG

is expressed in mg/dl (0.0551mmol/L) and PI in U/ml (6pmol/L). The early insulin secretion

index (∆PI30-0/∆PG30-0) is being calculated as the change in PI from 0 minutes to 30 minutes

divided by the change in PG over the same period [214].

5.4.13.3 Biomarkers of Adherence

The detection of SSB and water intake through 13C/12C ratios in serum fatty acids (e.g.

palmitoleic acid) derived from the de novo lipogenesis (DNL) of added sugars from sugar cane

(sucrose) or corn (high fructose corn syrup [HFCS]) is being assessed using a novel gas

chromatography, compound specific isotope ratio mass spectrometry (GC-CSIRMS) method at

the Department of Nutritional Sciences, University of Toronto. SSB and water intake adherence

is also being verified through an increase or decrease, respectively, of urinary fructose as

measured by gas chromatography mass spectrometry (GS-MS) [215]. NSB adherence is being

determined in urinary sucralose or Ace-K excretion through liquid chromatography with mass

spectrometry using electrospray ionization (LC-ESI-MS/MS) [216].

5.4.14 Antibiotic Use

Antibiotic use during an intervention phase is going to require a 30-day washout period at the

completion of the antibiotic course and a restart of that intervention phase. If the antibiotic

course is being taken during the washout phase, a 30-day additional washout is going to be

instituted.

112

5.4.15 Compensation

Participants are being provided monetary compensation for each study visit at the end of each

intervention phase. If a participant withdraws from the study, compensation is only being

provided for each visit that was completed. Participants are also being reimbursed for any study-

related transportation costs.

5.4.16 Statistical Analyses

An intention to treat principle using repeated measures mixed effects models in STATA 14

(StataCorp, Texas, USA) is being used to analyze the data. Sensitivity analysis is being

performed on the basis of complete data availability for primary endpoints. A

separate sensitivity analysis is being performed on the basis of antibiotic use during the trial.

For all primary and secondary outcomes effect modification by sex is being explored. The

truncated Benjamini-Hochberg false discovery rate controlling method with parallel gatekeeping

procedure is being used to correct for multiple comparisons for all primary and secondary

outcomes if at least one primary outcome reaches significance. If none of the primary outcomes

reach significance, the secondary outcomes are going to be analyzed as exploratory variables

with no adjustment for false discovery rate [201].

5.4.16.1 Subgroup Analysis

A priori subgroup analysis is being conducted by age, sex, ethnicity, baseline BMI, baseline

waist circumference, baseline FPG, baseline 2h-PG (75g OGTT), baseline iAUC, medication

use, sweeteners consumed from study beverages in the NSB arm, and background sweetener use.

5.4.17 Adverse Effects

Participants are being requested to report any adverse effects of the intervention (study

beverages) or visit examinations (75g OGTT solution, blood draw, stool sample collection)

immediately to study staff. No adverse effects are anticipated from the study drinks. Participants

may experience nausea after consuming the glucose drink, anxiety over having their blood drawn

and possible contamination of skin, clothing or food from the stool sample collection. To

mitigate these risks, participants are being given detailed instructions on the OGTT and blood

draw and proper collection of all bodily samples.

113

5.5 Results

5.5.1 Participant Flow

During the 18-month recruitment period, a total of 1086 individuals contacted the site to inquire

about the study. Of these individuals, 260 passed the telephone screen and were scheduled to

come in for the consent review. Main reasons for ineligibility from the phone screen included

consuming <1 355mL can SSB per day (n=54); low BMI (n=50); self-reported disease or illness

(n=48); did not have a primary care physician (n=28); or were smokers, recreational drug users

or heavy alcohol users (>3 drinks/day) (n=24). A total of 155 individuals consented and

completed the in-person screen, of which 141 were eligible and scheduled for their first visit.

Due to study withdrawal (n=36) and lost-to-follow up (n=22) a total of 81 participants were

randomized after completing visit one. A diagram of the flow of participants from enrollment to

randomization, following the Consolidated Standards of Reporting Trials (CONSORT) [217] is

shown in Figure 5.2.

5.5.2 Baseline Characteristics

Baseline characteristics of randomized participants in the STOP Sugars NOW trial are displayed

in Table 5.6. Mean age is 42 years. Forty-eight percent of participants are male and 52% are

female. Participants are predominately mixed (23.5%) and European (19.8%) decent with over

half born in Canada (59.3%). Among those who immigrated, the average year of immigration

was 1996. Many participants work ≥32 hours per week (50.6%), have an undergraduate degree

(33.3%) and report themselves as non-alcohol consumers (25.9%). Although subjects are mainly

obese (BMI 33.8kg/m2) with an elevated waist circumference (108.7cm), the mean blood

pressure level is normal (116.1/75.9), with only 6 participants (7%) on blood pressure lowering

medications. SSB intake (collected during the run-in phase), revealed Coke (43.2%) as the most

popular beverage consumed, followed by Canada Dry Ginger Ale (25.9%), and an overall mean

intake of two 355mL cans per day. Projected NSB intake, therefore, indicates that an

overwhelming majority of participants are going to be consuming a blend of aspartame and Ace-

K (95%). Most subjects indicated they did not consume any NNSs (60.5%), over the past 6

months, at their first visit. Of those who did, the most common source of NNS intake was from

NSBs (24.7%).

114

5.6 Discussion

The STOP Sugars NOW trial successfully recruited and enrolled 81 participants who are

receiving SSBs, NSBs and water in randomized order over the course of 4-weeks each. This

study is the first RCT to investigate how NSBs compare to water as a substitution for SSBs in

their effect on diversity of the gut microbiome and glucose tolerance.

This study overcomes several limitations of previous studies in the substitution of NSBs for

SSBs. It includes a “real-world” approach by assessing intake of NSBs based on products

available in the market as opposed to the addition of NNSs to beverages. Most NSBs are sold as

blends of NNSs to enhance the palatability of the product [91]. Therefore, adding them to

beverages as single-type sweeteners, often in amounts higher than those found on store shelfs,

prevents conclusions to be made on their efficacy as SSB replacements based on “real world”

intakes. Several studies have also failed to track the type and quantity of NNSs consumed in

beverages or investigate NNS intake through foods and beverages [156, 161-164, 168]. By

limiting NNS intake to NSBs and tracking intake through beverage logs, returned cans and

urinary biomarkers, possible deleterious (or beneficial) effects of NNS types or quantities can be

determined. Comparing NSBs for SSBs allows for the displacement of energy by the

comparator, which is often overlooked in RCT syntheses, including a recent WHO-

commissioned review [21]. Comparing NSBs for water (the “standard of care” replacement

beverage for SSBs), is going to clarify if NSBs are like water in their effect on gut bacteria and

diabetes risk.

Despite the strengths of this trial, there are challenges and limitations. Study length may decrease

adherence as the use of a crossover design requires participant involvement for a minimum of 5.5

months. Issues around antibiotic use; scheduling due to vacations; religious customs (fasting);

and changes in employment, school and family dynamics have arisen. These conflicts often lead

to an increase in study length through the extension of a washout phase to help participants

adhere to consuming their study beverages. Although our study population includes individuals

at risk for T2DM, we do not know if those living with the disease are going to see improvement

in glucose control when replacing their usual SSB intake with NSBs at “real-world” intakes.

Furthermore, even though we are assessing dietary intake through one of the best available

115

methods for measuring “free-living” food intake (weighted 3DDRs), it is possible that dietary

compensation resulting from displaced energy intake through NSBs or water, may not be

captured by these records, as dietary intake is not being assessed at the beginning of the 2nd and

3rd intervention phases.

In addition to this trial, a few others are currently investigating the effect of NNSs on outcomes

of cardiometabolic risk and/or gut microbiome. These trials include the “Microbiome and Non-

caloric Artificial Sweeteners in Humans” trial (ClinicalTrials.gov, NCT03708939), in which

participants are receiving NNSs compared to glucose on outcomes of glycemic control and gut

microbiome; the “Effects of Non-nutritive Sweeteners on the Composition of the Gut

Microbiome” trial (ClinicalTrials.gov, NCT02877186) in which participants are receiving NSBs

compared to water on outcomes of gut microbiome; and the “Effect of Non-nutritive Sweeteners

of High Sugar Sweetened Beverages on Metabolic Health and Gut Microbiome” trial

(ClinicalTrials.gov, NCT03259685) in which participants are receiving NSBs compared to SSBs

on outcomes of metabolic syndrome parameters and gut microbiome. Although these trials are

going to help strengthen the evidence base on the effect of NNSs on cardiometabolic risk and gut

microbiome, they do not assess the effect of NNSs through all 3 of our prespecified substitutions:

intended substitution with caloric displacement, “standard of care” substitution with caloric

displacement, and as a matched substitution without caloric displacement.

5.7 Conclusion

Although intakes of added sugars are declining in North America, they are still being consumed

above targets with SSBs being the most important source of added sugars [66, 218]. As a result,

strategies targeting the substitution of SSBs is an essential recommendation by authoritative

bodies. NSBs provide a viable alternative, yet their recommendation in dietary and clinical

practice guidelines are inconsistent [11-13, 15-18, 142], and concerns exists that they adversely

affect glucose tolerance by altering the composition of the gut microbiome [80, 81]. This unique

trial is going to inform guidelines on the usefulness of NSBs as SSB replacements; aid in

knowledge translation related to the health effects of NSBs for SSBs; and improve health

outcomes by educating healthcare providers and patients, stimulating industry innovation, and

guiding future research design.

116

5.8 Acknowledgements and Funding

A sincere thanks to all study participants; student volunteers; CNRFMC colleagues, including

the SMH Core Lab and MRS Research Centre, BBDC and the Department of Nutritional

Sciences (University of Toronto) who are making this study possible.

This work is being supported by the Canadian Institutes of Health Research (funding reference

number, 129920) through the Canada-wide Human Nutrition Trialists’ Network (NTN). The

Diet, Digestive tract, and Disease (3-D) Centre, funded through the Canada Foundation for

Innovation (CFI) and the Ministry of Research and Innovation’s Ontario Research Fund (ORF),

is providing the infrastructure for the conduct of this project. NM and SAC are being supported

by the Toronto 3D Internship and CIHR Canada Graduate Scholarship - Master's (CGS-M)

awards. NM is also receiving funding through the St. Michael’s Hospital Research Training

Centre Scholarship. JLS was funded by a PSI Graham Farquharson Knowledge Translation

Fellowship, Diabetes Canada Clinician Scientist award, CIHR INMD/CNS New Investigator

Partnership Prize, and Banting & Best Diabetes Centre Sun Life Financial New Investigator

Award. EMC holds the Lawson Family Chair in Microbiome Nutrition Research at the

University of Toronto. None of the sponsors have a role in any aspect of the present study,

including the design and conduct of the study; collection, management, analysis, and

interpretation of the data; and the preparation, review and approval of the manuscript or decision

to publish

5.9 Statement of Contribution

Néma McGlynn contributed to ethic submissions; provided input on how to build the trial

database in REDCap and participant recruitment; secured and trained study volunteers;

performed telephone and in-person screens; assisted in securing study materials; organized the

process for the purchase and delivery of study beverages; was a key player in the management

and execution of participant study visits; organized and managed the collection and storage of

participant data; helped set up a system for participant reminders; and performed all of the data

cleaning and analysis of baseline characteristics.

117

Visits

Visits

Visits

Visits

1Figure 5.1: Study Design1

NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages 1All possible groups not shown.

Anal4

1

5

3

S2

B

6

4

Group 1

Group 2

NSB

Group 3

SSB

NSB

Water Water

Washout

Washout

Washout

SSB

NSB

Water

Washout

Washout

Washout

SSB

NSB

Water

Randomization

4 weeks

≥4 weeks

4 weeks

≥4 weeks

4 weeks

164 Days (5.5 months)

164 Days (5.5 months)

Analysis

Analysis

1

1

5

5

3

3

2

2

6

6

s)

≥2 weeks

≥2 weeks

≥2 weeks

≥2 weeks

Run-in phase

Run-in phase

Run-in phase

Run-in phase

118

1Table 5.1: Study Exclusion Criteria1

Age between 18 and 75 years

BMI ≥23kg/m2 for Asian individuals and ≥25kg/m2 other individuals

Waist circumference ≥94cm in men, ≥80cm in women in Europid, Sub-Saharan African, Eastern Mediterranean, and

Middle Eastern individuals; ≥90cm in men and ≥80cm in women for South Asian, Chinese, Japanese, and South and

Central American individuals

Not regularly drinking SSBs (≥1 355mL can serving per day)2

Self-reported pregnant or breast-feeding females, or women planning on becoming pregnant throughout study duration

Regular medication use that have a clinically relevant effect on the primary outcomes (except birth control and PRN meds

such as Advil, Tylenol, etc.)

Antibiotic use in the last 3 months

Use of CAM deemed inappropriate by investigators

Self-reported diabetes

Self-reported hypertension (or systolic blood pressure ≥160mmHg or diastolic blood pressure ≥100mmHg [219])

Self-reported polycystic ovarian syndrome

Self-reported cardiovascular disease

Self-reported gastrointestinal disease

Previous bariatric surgery

Self-reported liver disease

Self-reported hyperthyroidism or hypothyroidism

Self-reported kidney disease

Self-reported chronic infection

Self-reported lung disease

Self-reported cancer/malignancy

Self-reported schizophrenia spectrum and other psychotic disorders, bipolar and related disorders, and dissociative disorders

Major surgery in the last 6 months

Other major illness or health-related incidence within the last 6 months

Smokes cigarettes

Regular recreational drug users

Heavy alcohol use (>3 drinks/day)

Does not have a primary care physician

Participation in any trials within the last 6 months or planning on participating in other trials for the duration of this study

Individuals planning on making dietary or physical activity changes throughout study duration

BMI, body mass index; CAM, complementary or alternative medicine; SSBs, sugar-sweetened beverages 1Disease exclusions will be based upon self-reported diagnosis. 2SSBs include sodas and soft drinks that contain at least 50 kcal per 8-oz serving. They do not include fruit drinks or

100% fruit juice, sports or energy drinks, sweetened iced tea or coffee, or homemade SSBs such as frescas or juices.

119

2Table 5.2 Power Calculation for Primary Outcomes Between Water and NSB Arms

Outcome

Type Outcome Mean Change SD Correlation N Alpha*

Power

(Beta)

N (corrected for

20% loss to follow

up)

Primary Outcomes Family

Primary

Microbiome

UniFrac

Distance

0.04 0.07 0.7 60 0.0375 98% 75

Primary Glucose iAUC

44.81

mmol/L/min

(20%)

113

mmol/L

/min

0.7 60 0.0375 89% 75

120

3Table 5.3: Latin Square Randomization for Three Group Crossover Study1

Intervention Phase 1 Intervention Phase 2 Intervention Phase 3

Sequence Group 1 A B C

Sequence Group 2 B C A

Sequence Group 3 C A B

Sequence Group 4 B A C

Sequence Group 5 A C B

Sequence Group 6 C B A 1Rows represent all possible sequences and columns represent different intervention phase.

4Table 5.4: Complete List of All Available Study Beverages

Ace-K, acesulfame potassium; NSBs, non-nutritive sweetened beverages; SSBs, sugar-sweetened beverages

SSBs group

355 ml can

42 grams of sugars

NSBs group

355 ml can

0 grams of sugar

Water group

355 ml bottle or

can

0 grams of sugar

Coca-Cola

Pepsi

Canada Dry Ginger

Ale

Schweppes Ginger

Ale

Mountain Dew

Sprite

7UP

Orange Crush

Diet Coke (Aspartame, Ace-K)

Coca Cola Zero (Aspartame, Ace-K)

Diet Pepsi (Aspartame, Ace-K)

Diet Canada Dry Ginger Ale (Aspartame, Ace-

K)

Diet Schweppes Ginger Ale (Ace-K, Sucralose)

Diet Mountain Dew (Aspartame, Ace-K,

Sucralose)

Sprite Zero (Aspartame, Ace-K)

Diet 7UP (Aspartame, Ace-K)

Diet Orange Crush (Sucralose)

Carbonated or

still water

121

5Table 5.5: Participant Visit Schedule

Phase1 Screening

(0 wks)

Run-in

(2 weeks)

Intervention

phase 1 (4 weeks)

Washout

(4 weeks)

Intervention

phase 2 (4 weeks)

Washout

(4 weeks)

Intervention

phase 3 (4 weeks)

Study Visit 0 No visit Week 0

(Start)

Week 4

(End)

No visit Week 0

(Start)

Week 4

(End)

No visit Week 0

(Start)

Week 4

(End)

Written

Informed

Consent

Provided

x

Anthropometric measures

Height x

Weight x x x x x x x

Waist

circumference

x x x x x x x

Blood pressure x x x x x x x

Biochemical measures

Fecal sample x x x x x x

24-hour urine

collection

(biomarkers

for adherence)

x x x x x x

Blood sample

drawn

x x x x x x

75g OGTT x x x x x x

Questionnaires

Personal

information

x

Demographic

Data

x

Medical

History

x

Case Report

Form (e.g.

medications,

physical

activity)

x x x x x x x

Beverage log x x x x x x x x x x

3DDR

(adherence to

background

diet)

x x x x

3DDR, 3-day diet record; OGTT, oral glucose tolerance test 1Length in weeks

122

2Figure 5.2: Trial Flow: Screening and Randomization

123

6Table 5.6 Baseline Characteristics of Randomized Participants1

Variable Value

Anthropometry Mean ± SD

Age (years) 41.8 ± 12.9

Females, n (%) 42 (51.9)

Height, cm 167.3 ± 10.6

Weight, kg 94.2 ± 18.9

BMI, kg/m2 33.8 ± 6.7

Waist circumference, cm 108.7 ± 13.4

Systolic blood pressure 116.1 ± 12.6

Diastolic blood pressure 75.9 ± 9.1

Resting pulse 72.9 ± 10.3

SSB preference n (mean intake/day)

7-UP 3 (2)

Coke 35 (2)

CD Ginger Ale 21 (1.5)

Orange Crush 3 (1)

Pepsi 11 (2)

Sprite 7 (2)

Unknown 1 (0)

Projected NSB intake (NNS blends) n (%)

Diet 7-UP (Asp & Ace-K) 3 (3.7)

Diet Coke (Asp & Ace-K) 16 (19.8)

Coke Zero (Asp & Ace-K) 19 (23.5)

CD Diet Ginger Ale (Asp & Ace-K) 21 (25.9)

Diet Orange Crush (Sucralose) 3 (3.7)

Diet Pepsi (Asp & Ace-K) 11 (13.6)

Sprite Zero (Asp & Ace-K) 7 (8.6)

Unknown (Unknown) 1 (1.2)

Non-nutritive sweetener intake n (%)

NSBs 20 (24.7)

NNS containing foods 3 (3.7)

Table-top NNS 9 (11.1)

None 49 (60.5)

Multiple categories selected 11 (13.6)

Ethnicity n (%)

Aboriginal 2 (2.5)

European 16 (19.8)

African/Caribbean 5 (6.2)

Scottish/Irish 7 (8.6)

English 5 (6.2)

Latin American 5 (6.2)

Indian 4 (4.9)

East Asian 6 (7.4)

South East Asian 6 (7.4)

124

Other 6 (7.4)

Mixed 19 (23.5)

Education level n (%)

High School 18 (22.2)

College or Diploma 18 (22.2)

Undergraduate Degree 27 (33.3)

Graduate Degree 16 (19.8)

Other 2 (2.5)

Work status n (%)

FT (≥ 32 hrs/wk) 41 (50.6)

PT (≤ 32 hrs/wk) 13 (16)

Casual 6 (7.4)

Stay at home parent 6 (7.4)

FT student 3 (3.7)

Disability 3 (3.7)

Other 9 (11.1)

Alcohol intake n (%)

None 21 (25.9)

1-2 times per year 9 (11.1)

Every 2-3 months 16 (19.8)

1-2 times per month 16 (19.8)

1-2 times per week 16 (19.8)

Daily 3 (3.7) Ace-K, acesulfame potassium; CD, Canada Dry; FT, full-time; NNS, non-nutritive sweetener; NSBs, non-

nutritive sweetened beverage; PT, part-time; SSB, sugar-sweetened beverage. 1N=81

125

Chapter 6: General Discussion

6.1 Summary

To determine whether non-nutritive sweetened beverages (NSBs) improve body weight and

cardiometabolic risk factors similar to water in their substitution for sugar-sweetened beverages

(SSBs), we conducted a systematic review and network meta-analysis of randomized controlled

trials (RCTs) assessing the effect of NSBs, as a replacement for SSBs, through 3 prespecified

substitutions: NSBs for SSBs (intended substitution with caloric displacement), water for SSBs

(“standard of care” substitution with caloric displacement) and NSBs for water (matched

substitution without caloric displacement) on measures of body weight and cardiometabolic risk

factors. Fourteen RCTs (n=1530), predominantly in people with overweight/obesity at risk for or

with diabetes, comparing NSBs, SSBs and/or water on the primary measure of body weight and

secondary outcomes of other measures of adiposity, glycemic control, blood lipids, blood

pressure, non-alcoholic fatty liver disease (NAFLD) and uric acid were identified. The

substitution of NSBs for SSBs improved body weight, body mass index (BMI), body fat,

triglycerides (TGs) and intra-hepatocellular lipid (IHCL). The substitution of water for SSBs

improved only uric acid and there was no effect of the substitution of NSBs for water except for

a small increase in hemoglobin A1c (HbA1c). The certainty of the evidence was moderate

(NSBs for SSBs) and low (water for SSBs and NSBs for water) for the primary outcome of body

weight and ranged from low to high for all other outcomes across all substitutions. Downgrades

were mainly made for imprecision. The available evidence supports the use of NSBs as an

alternative to water when used as a replacement for SSBs in overweight/obese people at risk for

or with diabetes. Due to imprecision in the estimate and concerns that NSBs adversely affect

glucose tolerance through compositional changes of the gut microbiome, there is a need for more

high-quality trials. To address this concern, we provided a design, rationale and baseline

characteristics report for a RCT investigating the effect of NSBs, as a replacement for SSBs,

through our 3 prespecified substitutions (NSBs for SSBs, water for SSBs and NSBs for water) on

gut microbiome and glucose tolerance over 4-weeks in overweight/obese participants who are

regular SSB drinkers: Strategies To OPpose SUGARS with Non-nutritive sweeteners Or Water

trial (STOP Sugars NOW). Baseline characteristics of study participants revealed an average

intake of two 355mL cans of SSBs per day, the majority of which were consuming Coke (n=37,

126

45.7%), followed by Canada Dry Ginger Ale (n=27, 33.3%). Projected NSB intake, based on the

brand-matched NSB options, indicates a NNS-blend of primarily aspartame and acesulfame

potassium (Ace-K) (n=77, 95%). This trial will be the first to determine if substituting NSBs for

SSBs, water for SSBs and NSBs for water impacts glucose control through changes in the gut

microbiota.

6.2 Strengths and Limitations

Strengths of the network meta-analysis are:

1. Accounting for the nature of the comparator to allow for the displacement of energy from

SSBs by NSBs in comparison to the preferred replacement beverage of water.

2. Use of network meta-analysis allowed for the quantitative comparison of our 3

prespecified beverage interventions that were not directly compared in trials. The

network meta-analysis then strengthened the evidence base by analyzing both direct and

indirect evidence collectively, thereby providing more precise estimates of the effect of

each beverage comparison on our outcomes.

3. A rigorous search and selection process of the available literature examining the effect of

NSBs, SSBs and water on outcomes of body weight and cardiometabolic risk factors.

4. Inclusion of only RCTs, a design that provides the greatest protection against bias.

5. Utilization of the GRADE approach to assess the certainty of our estimates.

Strengths of the STOP Sugars NOW trial are:

1. Implementation of the Benjamini-Hochberg procedure for the power calculation

which accounts for the adjustment required for multiple testing of two primary

outcomes (glucose iAUC and Microbiome UniFrac Distance).

2. Use of a “real-world” approach by assessing intake of NSBs based on products

available in the market, and through the equal displacement of SSBs with NSBs and

water based on each participant’s average intake at baseline.

3. Limiting NNS intake to NSBs and tracking adherence through urinary biomarkers to

determine possible deleterious (or beneficial) effects of NNS types or quantities.

127

4. Comparing NSBs to water in their substitution for SSBs which allows for the

displacement of energy by the comparator. This comparison will clarify if NSBs are

like water in their effect on gut bacteria and diabetes risk.

5. Although we will not be able to determine the effect of single NNSs on gut

microbiota changes and glucose tolerance, we are taking a pragmatic approach by

looking at “real-world” intakes of NNSs from NSBs, of which the majority exist as

blends of different NNSs.

Limitations of the network meta-analysis are:

1. Downgrades for evidence of serious inconsistency on the secondary adiposity outcome of

waist circumference in the substitution of NSBs for SSBs; on the secondary

cardiometabolic outcomes of HbA1c and FPI in the substitution of water for SSBs; and in

the primary outcome of body weight and several secondary adiposity and

cardiometabolic outcomes in the substitution of NSBs for water (waist circumference,

HbA1c, FPI, HOMA-IR, Non-HDL-C and TGs).

2. Downgrades for evidence of serious indirectness, as only one RCT or less of direct

comparisons was available for several secondary adiposity and cardiometabolic outcomes

in the analyses for the substitution of NSBs for SSBs (waist circumference, HbA1c, 2h-

PG, ALT and AST), water for SSBs (waist circumference, HbA1c, 2h-PG, FPI, HOMA-

IR, IHCL, ALT, AST and uric acid) and NSBs for water (IHCL, ALT, AST and uric

acid).

3. Downgrades for evidence of serious imprecision in the primary outcome of body weight

and several secondary adiposity and cardiometabolic outcomes in the substitution of

NSBs for SSBs (BMI, waist circumference, HbA1c, 2h-PG, FPI, LDL-C, Non-HDL-C,

TGs, HDL-C, TC, SBP, DBP, IHCL, ALT, AST and uric acid); water for SSBs (body

weight, BMI, waist circumference, HbA1c, 2h-PG, FPI, LDL-C, Non-HDL-C, TGs,

HDL-C, TC, SBP, DBP, IHCL, ALT, AST and uric acid); and NSBs for water (body

weight, BMI, waist circumference, HbA1c, 2h-PG, FPI, LDL-C, Non-HDL-C, TGs, TC,

SBP, DBP, IHCL, ALT, AST and uric acid) crossed our prespecified minimally

important differences. Accordingly, we downgraded the evidence in all cases for serious

imprecision.

128

4. The inability to assess for publication bias or conduct subgroup analyses as there were

less than 10 direct RCT comparisons per outcome for each of the 3 prespecified

substitutions. We were also unable to conduct sensitivity analyses in the network as

several of the RCTs contained multiple comparison arms.

Limitations/challenges of the STOP Sugars NOW trial are:

1. The length of the trial may decrease adherence as the use of a crossover design requires

participant involvement for a minimum of 5.5 months, which may further increase due to

each participants life’s circumstances (vacations; religious customs (fasting); and changes

in employment, school and family dynamic) and possible antibiotic use.

2. Our study population only includes individuals at risk for T2DM. Therefore, we won’t

know if those living with the disease will see improvement in glucose control or

microbiome diversity when replacing their usual SSB intake with NSBs at “real-world”

intakes.

3. Dietary compensation resulting from displaced energy intake through NSBs or water may

not be captured by weighted 3-day diet records (3DDRs), as dietary intake is not being

assessed at every study visit (only visits 1, 2, 4 and 6).

6.3 Clinical Implications

Current dietary and clinical practice guidelines are inconsistent in their recommendations for

NSBs [11-13, 15, 16, 18], even though results from SRMAs of RCTs consistently show non-

significant or beneficial effects of NSBs on outcomes of cardiometabolic risk in individuals with,

and without diabetes [19-21, 23-26, 138]. In our network meta-analysis, it was demonstrated that

when NSBs are substituted for SSBs, improvements are seen on body weight and

cardiometabolic outcomes over the moderate term with no evidence of harm. Although we did

not see benefits with the substitution of NSBs for SSBs on measures of glucose control, the

benefits observed on the primary outcome of body weight and other secondary outcomes of

adiposity, lipids and IHCL suggest that longer-term substitutions of NSBs for SSBs lead to

improvements in glycemic response with subsequent reduction in, and management of, diabetes

risk. Even though the findings from this NMA are limited to NSBs, they are of relevance to

NNSs in general as NNSs are added to several foods and beverages by consumers and product

129

developers in an attempt to reduce calories [85]. Most importantly, of all product categories

containing NNSs as a single food matrix, NSBs are globally the most consumed source [85].

Given the current concern with dietary sugars in the obesity and diabetes epidemic, the clinical

implications of our findings indicate that NSBs are a viable alternative to SSBs and are

comparable to water in their effect on body weight, glycemic control and cardiometabolic risk

factors. The use of NSBs as a displacement of energy from SSBs, however, should be part of a

broader strategy where guidance is given on replacements for all sources of excess caloric intake

from sugars. Emphasis should be placed on the overall quality of the diet by the incorporation of

a dietary pattern that best aligns with an individual’s values and preferences to help strengthen

adherence over the long-term [12].

6.4 Future Directions

The findings of this network meta-analysis are highly relevant for informing guidance on the role

of NSBs in the context of excess sugar intake from SSBs. While our results demonstrate that

NSBs have no adverse effect on blood glucose and insulin regulation in adults living with and

without diabetes, we were unable to establish any benefit on these outcomes when NSBs

substituted SSBs. Furthermore, as imprecision was detected in the network estimate for HbA1c,

2h-PG, and fasting insulin among all three beverage comparisons, there is a need for larger and

longer RCTs of a year or more to elucidate the potential benefits of NSBs on glycemic control.

Our inability to assess for publication bias or do subgroup analyses as there were less than 10

direct RCT comparisons per outcome for each of the 3 prespecified substitutions, and the sparse

network meta-analysis for outcomes of IHCL, ALT, AST and uric acid, resulting in downgrades

for serious imprecision among all 3 beverage comparisons further emphasize the need for more

high-quality RCTs. With additional RCTs we would be able to determine if the effect sizes

observed were overestimated and with false-positive results; if effect modifiers such as weight,

age, health status, background diet or type of NNS in the beverages could have influenced the

results; and if the results for outcomes with limited studies in the network were spurious.

The STOP Sugars NOW trial will not only add to the evidence-base on the effect of our 3

prespecified beverages on metabolic outcomes but will add to the limited evidence concerning

NNSs on glucose tolerance through possible changes in the gut microbiome based on “real-

world” intakes. If changes in the gut microbiome are observed with subsequent glucose

130

intolerance, additional well-designed RCTs will be required to elucidate a causal biological

plausibility of this relationship, while considering the nature of the comparator.

Considering 95% of participants are projected to consume a blend of aspartame and Ace-K,

determinations could also be made on the potential benefits or harms of this sweetener blend in

the NSBs consumed. Future trials could investigate the impact of these NNSs in certain foods or

other sweeteners that are gaining in popularity such as stevia. As the glucose units of steviol

glycosides, the sweet tasting compounds of the stevia plant, are cleaved off by bacteria in the

colon [91], a future study could mimic our trial design to investigate any potential effects this

NNS may have on glucose tolerance through changes in the gut microbiome, thereby adding to

the limited evidence of this sweetener on gut health [130].

Chapter 7: Conclusions

Conclusions:

The findings from this thesis demonstrated the following:

1. In a systematic review and network meta-analysis of 14 randomized controlled trials of

21 trial comparisons involving 1530 predominantly overweight or obese adult

participants, at risk for or with diabetes, over the moderate term, showed that substituting

non-nutritive sweetened beverages (NSBs) for sugar-sweetened beverages (SSBs), water

for SSBs and NSBs for water did not adversely affect cardiometabolic risk. Benefits were

observed for outcomes of body weight, BMI, body fat, triglycerides, and

intrahepatocellular lipids with the substitution of NSBs for SSBs. The overall certainty of

the evidence was rated as low for the substitutions of water for SSBs and NSBs for water,

and moderate for the substitution of NSBs for SSBs, with downgrades mainly for

imprecision, indicating a need for more high-quality trials to improve our confidence in

the estimates of effect.

2. The STOP Sugars NOW trial successful recruited 81 participants to receive SSBs, NSBs

and water in randomized order over the course of 4-weeks each. Baseline characteristics

indicated that participants consumed an average of two 355mL cans of SSBs per day with

131

a projected NNS intake of a blend of aspartame and Ace-K. This trial will address the

limitations of the trials used in our network meta-analysis while adding the necessary

evidence to improve our confidence in the estimates of effect of our 3 prespecified

substitutions (NSBs for SSBs, water for SSBs and NSBs for water) on body weight and

cardiometabolic risk. It will also help determine if NSBs for SSBs, water for SSBs and

NSBs for water, negatively affect the gut microbiome with associated developments of

adverse glucose tolerance.

This is the first network meta-analysis to investigate the role of NSBs compared to water as a

displacement of energy for SSBs on outcomes of body weight and cardiometabolic risk factors.

Our results are consistent with the previous meta-analyses of RCTs assessing the effect of NNSs

on body weight and cardiometabolic risk. The STOP Sugars NOW trial is the first RCT to

investigate how NSBs compare to water as a substitution for SSBs in their effect on diversity of

gut microbiome and glucose tolerance.

132

References

1. World Health Organization. Obesity and overweight. 2018; Available from: https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight.

2. Health Canada. Tackling obesity in canada: obesity and excess weight rates in canadian adults. 2018; Available from: https://www.canada.ca/en/public-health/services/publications/healthy-living/obesity-excess-weight-rates-canadian-adults.html.

3. O’Neill, M., K. Kornas, and L. Rosella, The future burden of obesity in Canada: a modelling study. Canadian Journal of Public Health, 2019. 110(6): p. 768-778.

4. The GBD 2015 Obesity Collaborators, Health effects of overweight and obesity in 195 countries over 25 years. New England Journal of Medicine, 2017. 377(1): p. 13-27.

5. World Health Organization. Global report on diabetes. 2016; Available from: https://apps.who.int/iris/bitstream/handle/10665/204871/9789241565257_eng.pdf;jsessionid=BD0EB38572D0002E58B5706B4FCF2E67?sequence=1#targetText=GlOBAl%20BurDEN,8.5%25%20in%20the%20adult%20population.

6. Malik, V.S., et al., Sugar-sweetened beverages and weight gain in children and adults: a systematic review and meta-analysis. Am J Clin Nutr, 2013. 98(4): p. 1084-102.

7. Imamura, F., et al., Consumption of sugar sweetened beverages, artificially sweetened beverages, and fruit juice and incidence of type 2 diabetes: systematic review, meta-analysis, and estimation of population attributable fraction. BMJ, 2015: p. h3576.

8. Jayalath, V.H., et al., Sugar-sweetened beverage consumption and incident hypertension: a systematic review and meta-analysis of prospective cohorts. Am J Clin Nutr, 2015. 102(4): p. 914-21.

9. Xi, B., et al., Sugar-sweetened beverages and risk of hypertension and CVD: a dose-response meta-analysis. Br J Nutr, 2015. 113(5): p. 709-17.

10. World Health Organization. Guideline: Sugars intake for adults and children. 2015; Available from: https://apps.who.int/iris/bitstream/handle/10665/149782/9789241549028_eng.pdf?sequence=1.

11. Health Canada. Canada's dietary guidelines for health professionals and policy makers. 2018; Available from: https://food-guide.canada.ca/static/assets/pdf/CDG-EN-2018.pdf.

12. Diabetes Canada, Clinical practice guidelines committees. Canadian Journal of Diabetes, 2018. 42.

13. United States Department of Agriculture. Scientific report of the 2015 dietary guidelines advisory committee. 2015.

14. Australian Government, Australian dietary guidelines, D.o.H.a. Ageing, Editor. 2013. 15. American Diabetes Association. Standards of medical care in diabetes - 2020. The journal of

Clinical and Applied Research and Education 2020; Available from: https://care.diabetesjournals.org/content/diacare/suppl/2019/12/20/43.Supplement_1.DC1/Standards_of_Care_2020.pdf.

16. Diabetes UK. Evidence-based nutrition guidelines for the prevention and management of diabetes. 2018; Available from: https://diabetes-resources-production.s3.eu-west-1.amazonaws.com/resources-s3/2018-03/1373_Nutrition%20guidelines_0.pdf.

17. Mann, J.I., et al., Evidence-based nutritional approaches to the treatment and prevention of diabetes mellitus. Nutr Metab Cardiovasc Dis, 2004. 14(6): p. 373-394.

18. Johnson, R.K., et al., Low-calorie sweetened beverages and cardiometabolic health: A science advisory from the american heart association. Circulation, 2018. 138(9).

https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight

https://www.canada.ca/en/public-health/services/publications/healthy-living/obesity-excess-weight-rates-canadian-adults.html

https://www.canada.ca/en/public-health/services/publications/healthy-living/obesity-excess-weight-rates-canadian-adults.html

https://apps.who.int/iris/bitstream/handle/10665/204871/9789241565257_eng.pdf;jsessionid=BD0EB38572D0002E58B5706B4FCF2E67?sequence=1#targetText=GlOBAl%20BurDEN,8.5%25%20in%20the%20adult%20population



https://apps.who.int/iris/bitstream/handle/10665/149782/9789241549028_eng.pdf?sequence=1

https://apps.who.int/iris/bitstream/handle/10665/149782/9789241549028_eng.pdf?sequence=1

https://food-guide.canada.ca/static/assets/pdf/CDG-EN-2018.pdf

https://care.diabetesjournals.org/content/diacare/suppl/2019/12/20/43.Supplement_1.DC1/Standards_of_Care_2020.pdf

https://care.diabetesjournals.org/content/diacare/suppl/2019/12/20/43.Supplement_1.DC1/Standards_of_Care_2020.pdf

https://diabetes-resources-production.s3.eu-west-1.amazonaws.com/resources-s3/2018-03/1373_Nutrition%20guidelines_0.pdf

https://diabetes-resources-production.s3.eu-west-1.amazonaws.com/resources-s3/2018-03/1373_Nutrition%20guidelines_0.pdf

133

19. Azad, M.B., et al., Nonnutritive sweeteners and cardiometabolic health: a systematic review and meta-analysis of randomized controlled trials and prospective cohort studies. Canadian Medical Association Journal, 2017. 189(28): p. E929-E939.

20. Rogers, P.J., et al., Does low-energy sweetener consumption affect energy intake and body weight? A systematic review, including meta-analyses, of the evidence from human and animal studies. International Journal of Obesity, 2016. 40(3): p. 381-394.

21. Toews, I., et al., Association between intake of non-sugar sweeteners and health outcomes: systematic review and meta-analyses of randomised and non-randomised controlled trials and observational studies. BMJ, 2019. 364: p. k4718.

22. Greenwood, D.C., et al., Association between sugar-sweetened and artificially sweetened soft drinks and type 2 diabetes: systematic review and dose–response meta-analysis of prospective studies. British Journal of Nutrition, 2014. 112(5): p. 725-734.

23. Miller, P.E. and V. Perez, Low-calorie sweeteners and body weight and composition: a meta-analysis of randomized controlled trials and prospective cohort studies. The American Journal of Clinical Nutrition, 2014. 100(3): p. 765-777.

24. Santos, N.C., et al., Metabolic effects of aspartame in adulthood: A systematic review and meta-analysis of randomized clinical trials. Crit Rev Food Sci Nutr, 2018. 58(12): p. 2068-2081.

25. Laviada-Molina, H., et al., Effects of nonnutritive sweeteners on body weight and BMI in diverse clinical contexts: Systematic review and meta-analysis. Obesity Reviews, 2020.

26. Onakpoya, I.J. and C.J. Heneghan, Effect of the natural sweetener, steviol glycoside, on cardiovascular risk factors: A systematic review and meta-analysis of randomised clinical trials. European Journal of Preventive Cardiology, 2015. 22(12): p. 1575-1587.

27. Obesity Canada. Report Card on Access to Obesity Treatment for Adults in Canada. . 2019; Available from: http://obesitycanada.ca/wp-content/uploads/2019/04/OC-Report-Card-2019-Eng-F-web.pdf.

28. Brauer, P., et al., Recommendations for prevention of weight gain and use of behavioural and pharmacologic interventions to manage overweight and obesity in adults in primary care. Canadian Medical Association Journal, 2015. 187(3): p. 184-195.

29. International Diabetes Federation. IDF diabetes atlas: Eighth edition. 2017; Available from: https://diabetesatlas.org/IDF_Diabetes_Atlas_8e_interactive_EN/.

30. Basu, S., et al., The relationship of sugar to population-level diabetes prevalence: an econometric analysis of repeated cross-sectional data. PLoS ONE, 2013. 8(2): p. e57873.

31. Tremmel, M., et al., Economic Burden of Obesity: A Systematic Literature Review. International Journal of Environmental Research and Public Health, 2017. 14(4): p. 435.

32. Bommer, C., et al., Global Economic Burden of Diabetes in Adults: Projections From 2015 to 2030. Diabetes Care, 2018. 41(5): p. 963-970.

33. Maluly, H.D.B., et al., Low- and No- Calorie Sweeteners (LNCS): critical evaluation of their safety and health risks. Food Science and Technology, 2020.

34. Leiter, L.A., et al., Cardiometabolic Risk in Canada: A Detailed Analysis and Position Paper by the Cardiometabolic Risk Working Group. Canadian Journal of Cardiology, 2011. 27(2): p. e1-e33.

35. Alberti, K.G.M.M., et al., Harmonizing the metabolic syndrome. Circulation, 2009. 120(16): p. 1640-1645.

36. Lau, D.C.W., et al., 2006 Canadian clinical practice guidelines on the management and prevention of obesity in adults and children Canadian Medical Association Journal, 2007. 176(8): p. 1-117.

37. Hsu, W.C., et al., BMI cut points to identify at-risk Asian Americans for type 2 diabetes screening. Diabetes Care, 2015. 38(1): p. 150-8.

http://obesitycanada.ca/wp-content/uploads/2019/04/OC-Report-Card-2019-Eng-F-web.pdf

http://obesitycanada.ca/wp-content/uploads/2019/04/OC-Report-Card-2019-Eng-F-web.pdf

https://diabetesatlas.org/IDF_Diabetes_Atlas_8e_interactive_EN/

134

38. WHO Expert Consultation, Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet, 2004. 363(9403): p. 157-63.

39. International Diabetes Federation. The IDF consensus worldwide definition of the metabolic syndrome. 2006; Available from: https://www.pitt.edu/~super1/Metabolic/IDF1.pdf.

40. Tan, C.E., et al., Can We Apply the National Cholesterol Education Program Adult Treatment Panel Definition of the Metabolic Syndrome to Asians? 2004. 27(5): p. 1182-1186.

41. World Health Organization, Obesity: Preventing and managing the global epidemic. WHO Technical Report Series, 2000. 894(i–xii): p. 1-253.

42. Adab, P., M. Pallan, and P.H. Whincup, Is BMI the best measure of obesity? BMJ, 2018: p. k1274. 43. Yusuf, S., et al., Obesity and the risk of myocardial infarction in 27,000 participants from 52

countries: a case-control study. Lancet, 2005. 366(9497): p. 1640-9. 44. World Health Organization, Waist circumference and waist–hip ratio: report of a WHO expert

consultation. 2008: Geneva, Switzerland. 45. Malik, V.S., et al., Sugar-sweetened beverages and risk of metabolic syndrome and type 2

diabetes: A meta-analysis. Diabetes Care, 2010. 33(11): p. 2477-2483. 46. Meng Wang, M.Y., Le Fang, Ru-Ying Hu, Association between sugar-sweetened beverages and

type 2 diabetes: A meta-analysis. Journal of Diabetes Investigation, 2015. 6(3): p. 360-366. 47. Malik, V.S., et al., Sugar-sweetened beverages, obesity, type 2 diabetes mellitus, and

cardiovascular disease risk. Circulation, 2010. 121(11): p. 1356-1364. 48. Te Morenga, L.A., et al., Dietary sugars and cardiometabolic risk: systematic review and meta-

analyses of randomized controlled trials of the effects on blood pressure and lipids. The American Journal of Clinical Nutrition, 2014. 100(1): p. 65-79.

49. Malik, V.S. and F.B. Hu, Sugar-sweetened beverages and cardiometabolic health: An update of the evidence. Nutrients, 2019. 11(8): p. 1840.

50. Dimeglio, D. and R. Mattes, Liquid versus solid carbohydrate: effects on food intake and body weight. International Journal of Obesity, 2000. 24(6): p. 794-800.

51. Cassady, B.A., R.V. Considine, and R.D. Mattes, Beverage consumption, appetite, and energy intake: what did you expect? The American Journal of Clinical Nutrition, 2012. 95(3): p. 587-593.

52. Mourao, D.M., et al., Effects of food form on appetite and energy intake in lean and obese young adults. International Journal of Obesity, 2007. 31(11): p. 1688-1695.

53. Tieken, S.M., et al., Effects of solid versus liquid meal-replacement products of similar energy content on hunger, satiety, and appetite-regulating hormones in older adults. Horm Metab Res, 2007. 39(5): p. 389-94.

54. Atkinson, F.S., K. Foster-Powell, and J.C. Brand-Miller, International tables of glycemic index and glycemic load values: 2008. Diabetes Care, 2008. 31(12): p. 2281-2283.

55. Tey, S.L., et al., Effects of aspartame-, monk fruit-, stevia- and sucrose-sweetened beverages on postprandial glucose, insulin and energy intake. International Journal of Obesity, 2017. 41(3): p. 450-457.

56. J Bloomer, R., et al., Blood glucose and insulin response to artificially- and sugar-sweetened sodas in healthy men. Integrative Food, Nutrition and Metabolism, 2016. 3(1).

57. Ludwig, D.S., The glycemic index. JAMA, 2002. 287(18): p. 2414. 58. Liu, S., et al., Relation between a diet with a high glycemic load and plasma concentrations of

high-sensitivity C-reactive protein in middle-aged women. The American Journal of Clinical Nutrition, 2002. 75(3): p. 492-498.

59. Chiavaroli, L., et al., Effect of fructose on established lipid targets: a systematic review and meta-analysis of controlled feeding trials. J Am Heart Assoc, 2015. 4(9): p. e001700.

60. David Wang, D., et al., Effect of fructose on postprandial triglycerides: a systematic review and meta-analysis of controlled feeding trials. Atherosclerosis, 2014. 232(1): p. 125-33.

https://www.pitt.edu/~super1/Metabolic/IDF1.pdf

135

61. Choo, V.L., et al., Food sources of fructose-containing sugars and glycaemic control: systematic review and meta-analysis of controlled intervention studies. BMJ, 2018: p. k4644.

62. Wang, D.D., et al., The effects of fructose intake on serum uric acid vary among controlled dietary trials. J Nutr, 2012. 142(5): p. 916-23.

63. Chiu, S., et al., Effect of fructose on markers of non-alcoholic fatty liver disease (NAFLD): a systematic review and meta-analysis of controlled feeding trials. European Journal of Clinical Nutrition, 2014. 68(4): p. 416-423.

64. Langlois, K., et al., Change in total sugars consumption among Canadian children and adults, Statistics Canada, Editor. January 16, 2019. p. 10-19.

65. Welsh, J.A., et al., Consumption of added sugars is decreasing in the United States. Am J Clin Nutr, 2011. 94(3): p. 726-34.

66. Bailey, R., et al., Sources of added sugars in young children, adolescents, and adults with low and high intakes of added sugars. Nutrients, 2018. 10(1): p. 102.

67. Rosinger, A., et al., Sugar-sweetened beverage consumption among U.S. adults, 2011–2014. NCHS Data Brief, 2017. 270: p. 1-8.

68. Singh, G.M., et al., Global, regional, and national consumption of sugar-sweetened beverages, fruit juices, and milk: A systematic assessment of Beverage intake in 187 countries. PLoS One, 2015. 10(8): p. e0124845.

69. Popkin, B.M. and C. Hawkes, Sweetening of the global diet, particularly beverages: patterns, trends, and policy responses. The Lancet Diabetes & Endocrinology, 2016. 4(2): p. 174-186.

70. Scientific Advisory Committee on Nutrition. Carbohydrates and Health. 2015; Available from: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/445503/SACN_Carbohydrates_and_Health.pdf.

71. Rippe, J.M., et al., What is the appropriate upper limit for added sugars consumption? Nutrition Reviews, 2017. 75(1): p. 18-36.

72. Colchero A, R.-D.J., Popkin BM, and Ng SW, In Mexico, evidence of sustained consumer response two years after implementing a sugar-sweetened beverage tax. Health Affairs, 2017. 36.

73. Diabetes Canada. Sugar & diabetes. 2019; Available from: https://www.diabetes.ca/en-CA/advocacy---policies/our-policy-positions/sugar---diabetes.

74. Public Health England. Sugar reduction: The evidence for action. 2015; Available from: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/470179/Sugar_reduction_The_evidence_for_action.pdf.

75. Mexican Academy of Medicine, Dietary and physical activity guidelines in the context of overweight and obesity in the Mexican population, Consejo Nacional de Ciencia y Tecnologia, Editor. 2015.

76. The Supreme Council of Health Health Promotion and Non-Communicable Diseases. Qatar dietary guidelines. 2015; Available from: http://www.fao.org/3/a-az908e.pdf.

77. Ministry of Health of Brazil. Dietary guidelines for the Brazil population. 2014; Available from: https://www.paho.org/hq/dmdocuments/2015/dietary-guides-brazil-eng.pdf.

78. Istituto Nazionale di Ricerca per gli Alimenti e la Nutrizione. Linee guida per una sana alimentazione italiana. 2003; Available from: http://www.fao.org/3/a-as686o.pdf.

79. Sylvetsky, A.C. and K.I. Rother, Nonnutritive sweeteners in weight management and chronic disease: A review. Obesity (Silver Spring), 2018. 26(4): p. 635-640.

80. Hunter, S.R., et al., Low Calorie Sweeteners Differ in Their Physiological Effects in Humans. Nutrients, 2019. 11(11): p. 2717.

81. Suez, J., et al., Artificial sweeteners induce glucose intolerance by altering the gut microbiota. Nature, 2014. 514(7521): p. 181-186.

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/445503/SACN_Carbohydrates_and_Health.pdf

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/445503/SACN_Carbohydrates_and_Health.pdf

https://www.diabetes.ca/en-CA/advocacy---policies/our-policy-positions/sugar---diabetes

https://www.diabetes.ca/en-CA/advocacy---policies/our-policy-positions/sugar---diabetes

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/470179/Sugar_reduction_The_evidence_for_action.pdf

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/470179/Sugar_reduction_The_evidence_for_action.pdf

http://www.fao.org/3/a-az908e.pdf

https://www.paho.org/hq/dmdocuments/2015/dietary-guides-brazil-eng.pdf

http://www.fao.org/3/a-as686o.pdf

136

82. Diabetes Canada. Sugars and sweeteners. 2018; Available from: https://guidelines.diabetes.ca/docs/patient-resources/sugars-and-sweeteners.pdf.

83. Health Canada. Sugar alcohols (polyols) and polydextrose used as sweeteners in foods 2018; Available from: https://www.canada.ca/en/health-canada/services/food-nutrition/food-safety/food-additives/sugar-substitutes/sugar-alcohols-polyols-polydextrose-used-sweeteners-foods-food-safety.html.

84. Health Canada. Sweeteners definitions. 2018; Available from: http://www.inspection.gc.ca/food/requirements-and-guidance/labelling/industry/sweeteners/eng/1387749708758/1387750396304?chap=5.

85. Sylvetsky, A.C. and K.I. Rother, Trends in the consumption of low-calorie sweeteners. Physiology & Behavior, 2016. 164: p. 446-450.

86. European Food Safety Authority. Sweeteners. Available from: http://www.efsa.europa.eu/en/topics/topic/sweeteners.

87. U.S. Food and Drug Administration. High-intensity sweeteners. 2017; Available from: https://www.fda.gov/food/food-additives-petitions/high-intensity-sweeteners.

88. Health Canada. The safety of sugar substitutes. 2008; Available from: https://www.canada.ca/en/health-canada/services/healthy-living/your-health/food-nutrition/safety-sugar-substitutes.html.

89. Serra-Majem, L., et al., Ibero–american consensus on low- and no-calorie sweeteners: Safety, nutritional aspects and benefits in food and beverages. Nutrients, 2018. 10(7): p. 818.

90. World Health Organization. Alphabetical list of WHO member states. [cited 2019; Available from: https://www.who.int/choice/demography/by_country/en/.

91. Magnuson, B.A., et al., Biological fate of low-calorie sweeteners. Nutrition Reviews, 2016. 74(11): p. 670-689.

92. Health Canada. List of permitted sweeteners (lists of permitted food additives). 2019; Available from: https://www.canada.ca/en/health-canada/services/food-nutrition/food-safety/food-additives/lists-permitted/9-sweeteners.html.

93. Ahmad, S.Y., J.K. Friel, and D.S. MacKay, The effect of the artificial sweeteners on glucose metabolism in healthy adults: a randomized double-blinded crossover clinical trial. Appl Physiol Nutr Metab, 2019.

94. Nettleton, J.E., R.A. Reimer, and J. Shearer, Reshaping the gut microbiota: Impact of low calorie sweeteners and the link to insulin resistance? Physiol Behav, 2016. 164(Pt B): p. 488-493.

95. Franz, M. Amounts of sweeteners in popular diet sodas. Diabetes Self-Management 2010; Available from: https://static.diabetesselfmanagement.com/pdfs/DSM0310_012.pdf

96. Calorie Control Council. Cyclamate information center. 2018; Available from: https://www.cyclamate.org/facts.html.

97. Markets, M.a. Sugar substitutes market by type, composition, application, & by region: Forecast to 2020. 2015; Available from: https://www.marketsandmarkets.com/Market-Reports/sugar-substitute-market-1134.html.

98. Magnuson, B.A., A. Roberts, and E.R. Nestmann, Critical review of the current literature on the safety of sucralose. Food and Chemical Toxicology, 2017. 106: p. 324-355.

99. Canadian Food Inspection Agency. Labelling requirements for sweeteners and foods that contain sweeteners. 2020; Available from: https://www.inspection.gc.ca/food-label-requirements/labelling/industry/sweeteners/eng/1387749708758/1387750396304?chap=3.

100. Burke, M.V. and D.M. Small, Physiological mechanisms by which non-nutritive sweeteners may impact body weight and metabolism. Physiology & Behavior, 2015. 152: p. 381-388.

101. Dhillon, J., J.Y. Lee, and R.D. Mattes, The cephalic phase insulin response to nutritive and low-calorie sweeteners in solid and beverage form. Physiology & Behavior, 2017. 181: p. 100-109.

https://guidelines.diabetes.ca/docs/patient-resources/sugars-and-sweeteners.pdf

https://www.canada.ca/en/health-canada/services/food-nutrition/food-safety/food-additives/sugar-substitutes/sugar-alcohols-polyols-polydextrose-used-sweeteners-foods-food-safety.html



http://www.inspection.gc.ca/food/requirements-and-guidance/labelling/industry/sweeteners/eng/1387749708758/1387750396304?chap=5

http://www.inspection.gc.ca/food/requirements-and-guidance/labelling/industry/sweeteners/eng/1387749708758/1387750396304?chap=5

http://www.efsa.europa.eu/en/topics/topic/sweeteners

https://www.fda.gov/food/food-additives-petitions/high-intensity-sweeteners

https://www.canada.ca/en/health-canada/services/healthy-living/your-health/food-nutrition/safety-sugar-substitutes.html

https://www.canada.ca/en/health-canada/services/healthy-living/your-health/food-nutrition/safety-sugar-substitutes.html

https://www.who.int/choice/demography/by_country/en/

https://www.canada.ca/en/health-canada/services/food-nutrition/food-safety/food-additives/lists-permitted/9-sweeteners.html

https://www.canada.ca/en/health-canada/services/food-nutrition/food-safety/food-additives/lists-permitted/9-sweeteners.html

https://static.diabetesselfmanagement.com/pdfs/DSM0310_012.pdf

https://www.cyclamate.org/facts.html

https://www.marketsandmarkets.com/Market-Reports/sugar-substitute-market-1134.html

https://www.marketsandmarkets.com/Market-Reports/sugar-substitute-market-1134.html

https://www.inspection.gc.ca/food-label-requirements/labelling/industry/sweeteners/eng/1387749708758/1387750396304?chap=3

https://www.inspection.gc.ca/food-label-requirements/labelling/industry/sweeteners/eng/1387749708758/1387750396304?chap=3

137

102. Just, T., et al., Cephalic phase insulin release in healthy humans after taste stimulation? 2008. 51(3): p. 622-627.

103. Swithers, S.E., A.A. Martin, and T.L. Davidson, High-intensity sweeteners and energy balance. Physiology & Behavior, 2010. 100(1): p. 55-62.

104. Tucker, R.M. and S.Y. Tan, Do non-nutritive sweeteners influence acute glucose homeostasis in humans? A systematic review. Physiol Behav, 2017. 182: p. 17-26.

105. Kojima, I., et al., Glucose-sensing receptor T1R3: A new signaling receptor activated by glucose in pancreatic β-cells. 2015. 38(5): p. 674-679.

106. Kyriazis, G.A., M.M. Soundarapandian, and B. Tyrberg, Sweet taste receptor signaling in beta cells mediates fructose-induced potentiation of glucose-stimulated insulin secretion. 2012. 109(8): p. E524-E532.

107. Nakagawa, Y., et al., Sweet taste receptor expressed in pancreatic β-cells activates the calcium and cyclic AMP signaling systems and stimulates insulin secretion. PLoS ONE, 2009. 4(4): p. e5106.

108. Margolskee, R.F., et al., T1R3 and gustducin in gut sense sugars to regulate expression of Na+-glucose cotransporter 1. Proceedings of the National Academy of Sciences, 2007. 104(38): p. 15075-15080.

109. Moran, A.W., et al., Expression of Na+/glucose co-transporter 1 (SGLT1) is enhanced by supplementation of the diet of weaning piglets with artificial sweeteners. British Journal of Nutrition, 2010. 104(05): p. 637-646.

110. Stearns, A.T., et al., Rapid upregulation of sodium-glucose transporter SGLT1 in response to intestinal sweet taste stimulation. Ann Surg, 2010. 251(5): p. 865-71.

111. Mace, O.J., et al., Sweet taste receptors in rat small intestine stimulate glucose absorption through apical GLUT2. The Journal of Physiology, 2007. 582(1): p. 379-392.

112. Jang, H.J., et al., Gut-expressed gustducin and taste receptors regulate secretion of glucagon-like peptide-1. Proceedings of the National Academy of Sciences, 2007. 104(38): p. 15069-15074.

113. Brown, R.J. and K.I. Rother, Non-nutritive sweeteners and their role in the gastrointestinal tract. The Journal of Clinical Endocrinology & Metabolism, 2012. 97(8): p. 2597-2605.

114. Brown, R.J., M. Walter, and K.I. Rother, Effects of diet soda on gut hormones in youths with diabetes. Diabetes Care, 2012. 35(5): p. 959-964.

115. Sylvetsky, A.C., et al., Hormonal responses to non-nutritive sweeteners in water and diet soda. Nutr Metab (Lond), 2016. 13(1): p. 71.

116. Pepino, M.Y., et al., Sucralose affects glycemic and hormonal responses to an oral glucose load. Diabetes Care, 2013. 36(9): p. 2530-2535.

117. Wittekind, A., et al., A workshop on "dietary sweetness-is it an issue?'. International Journal of Obesity, 2018. 42: p. 934-938.

118. Zhang, G.H., et al., Effects of mother's dietary exposure to acesulfame-K in Pregnancy or lactation on the adult offspring's sweet preference. Chem Senses, 2011. 36(9): p. 763-70.

119. Drewnowski, A. and C. Rehm, Consumption of low-calorie sweeteners among U.S. adults is associated with higher healthy eating index (HEI 2005) scores and more physical activity. 2014. 6(10): p. 4389-4403.

120. Gibson, S., et al., Low calorie beverage consumption is associated with energy and nutrient intakes and diet quality in british adults. Nutrients, 2016. 8(1): p. 9.

121. Patel, L., G. Alicandro, and C. La Vecchia, Low-calorie beverage consumption, diet quality and cardiometabolic risk factors in british adults. Nutrients, 2018. 10(9): p. 1261.

122. Leahy, M., et al., Consumption of low-calorie sweetened beverages compared to water is associated with reduced intake of carbohydrates and sugar, with no adverse relationships to

138

glycemic responses: Results from the 2001–2012 national health and nutrition examination survey. Nutrients, 2017. 9(9): p. 928.

123. Ashwell, M., et al., Expert consensus on low-calorie sweeteners: facts, research gaps and suggested actions. Nutrition Research Reviews, 2020: p. 1-10.

124. Abou-Donia, M.B., et al., Splenda alters gut microflora and increases intestinal p-glycoprotein and cytochrome p-450 in male rats. Journal of Toxicology and Environmental Health, Part A, 2008. 71(21): p. 1415-1429.

125. Uebanso, T., et al., Effects of low-dose non-caloric sweetener consumption on gut microbiota in mice. Nutrients, 2017. 9(6): p. 560.

126. Wang, Q.P., et al., Non-nutritive sweeteners possess a bacteriostatic effect and alter gut microbiota in mice. PLoS One, 2018. 13(7): p. e0199080.

127. Bian, X., et al., Saccharin induced liver inflammation in mice by altering the gut microbiota and its metabolic functions. Food and Chemical Toxicology, 2017. 107: p. 530-539.

128. Anderson, R.L. and J.J. Kirkland, The effect of sodium saccharin in the diet on caecal microflora. 1980. 18(4): p. 353-355.

129. Naim, M., et al., Effects of sodium saccharin on the activity of trypsin, chymotrypsin, and amylase and upon bacteria in small intestinal contents of rats. 1985. 178(3): p. 392-401.

130. Mela, D.J., J. McLaughlin, and P.J. Rogers, Perspective: Standards for research and reporting on low-energy (“artificial”) sweeteners. Advances in Nutrition, 2020.

131. Thomson, P., et al., Short-term impact of sucralose consumption on the metabolic response and gut microbiome of healthy adults. Br J Nutr, 2019. 122(8): p. 856-862.

132. Ruanpeng, D., et al., Sugar and artificially sweetened beverages linked to obesity: a systematic review and meta-analysis. QJM: An International Journal of Medicine, 2017. 110(8): p. 513-520.

133. Cheungpasitporn, W., et al., Associations of sugar-sweetened and artificially sweetened soda with chronic kidney disease: A systematic review and meta-analysis. Nephrology, 2014. 19(12): p. 791-797.

134. Malik, V.S., Non-sugar sweeteners and health. BMJ, 2019: p. k5005. 135. Pan, A., et al., Changes in water and beverage intake and long-term weight changes: results

from three prospective cohort studies. International Journal of Obesity, 2013. 37(10): p. 1378-1385.

136. Mozaffarian, D., et al., Changes in diet and lifestyle and long-term weight gain in women and men. New England Journal of Medicine, 2011. 364(25): p. 2392-2404.

137. Field-Fote, E., Mediators and Moderators, Confounders and Covariates. Journal of Neurologic Physical Therapy, 2019. 43(2): p. 83-84.

138. Wiebe, N., et al., A systematic review on the effect of sweeteners on glycemic response and clinically relevant outcomes. BMC Medicine, 2011. 9(1): p. 123.

139. Riley, R.D., et al., Multivariate and network meta-analysis of multiple outcomes and multiple treatments: rationale, concepts, and examples. BMJ, 2017: p. j3932.

140. Dias, S. and D.M. Caldwell, Network meta-analysis explained. Archives of Disease in Childhood - Fetal and Neonatal Edition, 2019. 104(1): p. F8-F12.

141. Tonin, F.S., et al., Network meta-analysis: a technique to gather evidence from direct and indirect comparisons. Pharmacy Practice, 2017. 15(1): p. 943-943.

142. Evert, A.B., et al., Nutrition therapy for adults with diabetes or prediabetes: a consensus report. Diabetes Care, 2019. 42(5): p. 731-754.

143. Sievenpiper, J.L., et al., The importance of study design in the assessment of nonnutritive sweeteners and cardiometabolic health. Canadian Medical Association Journal, 2017. 189(46): p. E1424-E1425.

139

144. Puhan, M.A., et al., A GRADE Working Group approach for rating the quality of treatment effect estimates from network meta-analysis. 2014. 349(sep24 5): p. g5630-g5630.

145. Higgins, J., Thomas, J. Cochrane handbook for systematic reviews of interventions. Ed 6. 2019; Available from: https://training.cochrane.org/handbook/current.

146. Hutton, B., et al., The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: Checklist and explanations. Annals of Internal Medicine, 2015. 162(11): p. 777.

147. Chaimani, A., et al., Graphical tools for network meta-analysis in stata. PLoS ONE, 2013. 8(10): p. e76654.

148. Dias, S., et al., Checking consistency in mixed treatment comparison meta-analysis. Statistics in Medicine, 2010. 29(7-8): p. 932-944.

149. Jackson, D., et al., A design-by-treatment interaction model for network meta-analysis with random inconsistency effects. 2014. 33(21): p. 3639-3654.

150. Guyatt, G., et al., GRADE guidelines: 1. Introduction—GRADE evidence profiles and summary of findings tables. Journal of Clinical Epidemiology, 2011. 64(4): p. 383-394.

151. Brignardello-Petersen, R., et al., Advances in the GRADE approach to rate the certainty in estimates from a network meta-analysis. Journal of Clinical Epidemiology, 2018. 93: p. 36-44.

152. Brignardello-Petersen, R., et al., GRADE approach to rate the certainty from a network meta-analysis: avoiding spurious judgments of imprecision in sparse networks. Journal of Clinical Epidemiology, 2019. 105: p. 60-67.

153. Brignardello-Petersen, R., et al., GRADE approach to rate the certainty from a network meta-analysis: addressing incoherence. Journal of Clinical Epidemiology, 2019. 108: p. 77-85.

154. Bonnet, F., et al., Consumption of a carbonated beverage with high-intensity sweeteners has no effect on insulin sensitivity and secretion in nondiabetic adults. The Journal of Nutrition, 2018. 148(8): p. 1293-1299.

155. Bruun, J.M., et al., Consumption of sucrose-sweetened soft drinks increases plasma levels of uric acid in overweight and obese subjects: a 6-month randomised controlled trial. European Journal of Clinical Nutrition, 2015. 69(8): p. 949-953.

156. Campos, V., et al., Sugar- and artificially sweetened beverages and intrahepatic fat: A randomized controlled trial. Obesity, 2015. 23(12): p. 2335-2339.

157. Engel, S., et al., Effect of high milk and sugar-sweetened and non-caloric soft drink intake on insulin sensitivity after 6 months in overweight and obese adults: a randomized controlled trial. European Journal of Clinical Nutrition, 2018. 72(3): p. 358-366.

158. Hernández-Cordero, S., et al., Substituting Water for Sugar-Sweetened Beverages Reduces Circulating Triglycerides and the Prevalence of Metabolic Syndrome in Obese but Not in Overweight Mexican Women in a Randomized Controlled Trial. The Journal of Nutrition, 2014. 144(11): p. 1742-1752.

159. Higgins, K.A., R.V. Considine, and R.D. Mattes, Aspartame consumption for 12 weeks does not affect glycemia, appetite, or body weight of healthy, lean adults in a randomized controlled trial. The Journal of Nutrition, 2018. 148(4): p. 650-657.

160. Higgins, K.A. and R.D. Mattes, A randomized controlled trial contrasting the effects of 4 low-calorie sweeteners and sucrose on body weight in adults with overweight or obesity. The American Journal of Clinical Nutrition, 2019. 109(5): p. 1288-1301.

161. Madjd, A., et al., Effects on weight loss in adults of replacing diet beverages with water during a hypoenergetic diet: a randomized, 24-wk clinical trial. The American Journal of Clinical Nutrition, 2015. 102(6): p. 1305-1312.

https://training.cochrane.org/handbook/current

140

162. Madjd, A., et al., Beneficial effects of replacing diet beverages with water on type 2 diabetic obese women following a hypo-energetic diet: A randomized, 24-week clinical trial. Diabetes, Obesity and Metabolism, 2017. 19(1): p. 125-132.

163. Maersk, M., et al., Sucrose-sweetened beverages increase fat storage in the liver, muscle, and visceral fat depot: a 6-mo randomized intervention study. The American Journal of Clinical Nutrition, 2012. 95(2): p. 283-289.

164. Peters, J.C., et al., The effects of water and non-nutritive sweetened beverages on weight loss and weight maintenance: A randomized clinical trial. Obesity, 2016. 24(2): p. 297-304.

165. Reid, M., et al., Long-term dietary compensation for added sugar: effects of supplementary sucrose drinks over a 4-week period. British Journal of Nutrition, 2007. 97(1): p. 193-203.

166. Reid, M., R. Hammersley, and M. Duffy, Effects of sucrose drinks on macronutrient intake, body weight, and mood state in overweight women over 4 weeks. Appetite, 2010. 55(1): p. 130-136.

167. Reid, M., et al., Effects on obese women of the sugar sucrose added to the diet over 28 d: a quasi-randomised, single-blind, controlled trial. British Journal of Nutrition, 2014. 111(3): p. 563-570.

168. Tate, D.F., et al., Replacing caloric beverages with water or diet beverages for weight loss in adults: main results of the Choose Healthy Options Consciously Everyday (CHOICE) randomized clinical trial. Am J Clin Nutr, 2012. 95(3): p. 555-63.

169. Tordoff, M.G. and A.M. Alleva, Effect of drinking soda sweetened with aspartame or high-fructose corn syrup on food intake and body weight. Am J Clin Nutr, 1990. 51(6): p. 963-9.

170. Sievenpiper, J.L., et al., Effect of fructose on body weight in controlled feeding trials. 2012. 156(4): p. 291.

171. De Ruyter, J.C., et al., A trial of sugar-free or sugar-sweetened beverages and body weight in children. New England Journal of Medicine, 2012. 367(15): p. 1397-1406.

172. Hu, F.B., Globalization of Diabetes: The role of diet, lifestyle, and genes. Diabetes Care, 2011. 34(6): p. 1249-1257.

173. Lustig, R.H., L.A. Schmidt, and C.D. Brindis, The toxic truth about sugar. Nature, 2012. 482(7383): p. 27-29.

174. Kirkpatrick, S.I., et al., Top dietary sources of energy, sodium, sugars, and saturated fats among Canadians: insights from the 2015 Canadian Community Health Survey. Applied Physiology, Nutrition, and Metabolism, 2019. 44(6): p. 650-658.

175. Canadian Diabetes Association. Diabetes: Canada at the tipping point. Available from: file:///C:/Users/Student/Downloads/WEB-Eng-CDA-Report.pdf.

176. Canadian Medical Association Journal. Artificial sweeteners linked to risk of weight gain, heart disease and other health issues. 2017; Available from: https://www.sciencedaily.com/releases/2017/07/170717091043.htm.

177. Azad, M. The effects of nonnutritive sweeteners on cardiometabolic health. Physician's Weekly 2018; Available from: https://www.physiciansweekly.com/the-effects-of-nonnutritive-sweeteners-on-cardiometabolic-health/.

178. LaMotte, S. Artificial sweeteners: Where do we stand? CNN Health 2018; Available from: https://www.cnn.com/2016/01/18/health/where-do-we-stand-artificial-sweeteners/index.html.

179. Shreeves, R. What artificial sweeteners do to your body: They've been linked to diabetes, dementia and decreased gut health. MNN.com 2019; Available from: https://www.mnn.com/food/healthy-eating/blogs/scientists-may-have-found-link-between-artificial-sweeteners-and-diabetes.

180. Beck, L. Evidence suggesting artificial sweeteners may be harmful should give us pause. The Globe and Mail 2018; Available from: https://www.theglobeandmail.com/life/health-and-fitness/article-evidence-suggesting-artificial-sweeteners-may-be-harmful-should-

https://www.sciencedaily.com/releases/2017/07/170717091043.htm

https://www.physiciansweekly.com/the-effects-of-nonnutritive-sweeteners-on-cardiometabolic-health/

https://www.physiciansweekly.com/the-effects-of-nonnutritive-sweeteners-on-cardiometabolic-health/

https://www.cnn.com/2016/01/18/health/where-do-we-stand-artificial-sweeteners/index.html

https://www.mnn.com/food/healthy-eating/blogs/scientists-may-have-found-link-between-artificial-sweeteners-and-diabetes

https://www.mnn.com/food/healthy-eating/blogs/scientists-may-have-found-link-between-artificial-sweeteners-and-diabetes

https://www.theglobeandmail.com/life/health-and-fitness/article-evidence-suggesting-artificial-sweeteners-may-be-harmful-should-give/?utm_medium=Newsletter&utm_source=Globe%20Life&utm_type=text&utm_content=GlobeLife&utm_campaign=2018-10-15_16


141

give/?utm_medium=Newsletter&utm_source=Globe%20Life&utm_type=text&utm_content=GlobeLife&utm_campaign=2018-10-15_16.

181. Brueck, H. More evidence just landed that artificial sweeteners really are bad for you. Science Alert 2018; Available from: https://www.sciencealert.com/artificial-sweeteners-like-aspartame-can-be-toxic-to-our-gut-microbes.

182. Kirkey, S. Fake sugars linked to obesity, heart disease. National Post 2017; Available from: http://nationalpost.com/news/0717-na-sugar.

183. Leung, W. Sugar substitutes associated with weight gain and health problems, study says. Globe and Mail 2017; Available from: https://www.theglobeandmail.com/news/national/sugar-substitutes-linked-to-weight-gain-and-health-problems-study-says/article35704562/.

184. Gardener, H. and M.S.V. Elkind, Artificial sweeteners, real risks. Stroke, 2019. 50(3): p. 549-551. 185. Pagoto, S.L. and B.M. Appelhans, A Call for an end to the diet debates. JAMA, 2013. 310(7): p.

687-688. 186. Heart and Stroke Foundation of Canada. Sugar, heart disease and stroke. 2014; Available from:

https://www.heartandstroke.ca/-/media/pdf-files/canada/2017-position-statements/sugar-ps-eng.ashx.

187. US Department of Health and Human Services. Cut down on added sugars. 2016; Available from: https://health.gov/dietaryguidelines/2015/resources/DGA_Cut-Down-On-Added-Sugars.pdf.

188. Malik, V.S., et al., Long-term consumption of sugar-sweetened and artificially sweetened beverages and risk of mortality in US adults. Circulation, 2019. 139(18): p. 2113-2125.

189. Chang, K. Artificial sweeteners may disrupt body’s blood sugar controls. The New York Times 2014; Available from: https://well.blogs.nytimes.com/2014/09/17/artificial-sweeteners-may-disrupt-bodys-blood-sugar-controls/

190. Faherty, S. Artificial sweeteners may have despicable impacts on gut microbes. Scientific American 2014; Available from: https://blogs.scientificamerican.com/guest-blog/artificial-sweeteners-may-have-despicable-impacts-on-gut-microbes/

191. Sifferlin, A. Artificial sweeteners are linked to weight gain - not weight loss. Time 2017; Available from: http://time.com/4859012/artificial-sweeteners-weight-loss/.

192. Vergano, D. Artificial sweeteners may trigger blood sugar risks. National Geographic 2014; Available from: http://news.nationalgeographic.com/news/2014/09/140917-sweeteners-artificial-blood-sugar-diabetes-health-ngfood/

193. Brait, E. More research needed into the effects sugar substitutes have on health. The Toronto Star 2017; Available from: https://www.thestar.com/life/2017/07/17/more-research-needed-into-the-effects-sugar-substitutes-have-on-health.html

194. Thompson, D. Do artificial sweeteners raise odds for obesity? . WebMD 2017; Available from: http://www.webmd.com/food-recipes/news/20170717/do-artificial-sweeteners-raise-odds-for-obesity#1

195. Canadian Institutes of Health Research in partnership with Health Canada. Operating grant : Sugar and health ARCHIVED. 2016; Available from: https://www.researchnet-recherchenet.ca/rnr16/vwOpprtntyDtls.do?prog=2554&printfriendly=true.

196. Zeevi, D., et al., Personalized nutrition by prediction of glycemic responses. Cell, 2015. 163(5): p. 1079-1094.

197. Kelly, B.J., et al., Power and sample-size estimation for microbiome studies using pairwise distances and PERMANOVA. Bioinformatics, 2015. 31(15): p. 2461-2468.

198. David, L.A., et al., Diet rapidly and reproducibly alters the human gut microbiome. Nature, 2014. 505(7484): p. 559-563.

199. Braunstein, C., et al., A double-blind, randomized controlled, acute feeding equivalence trial of small, catalytic doses of fructose and allulose on postprandial blood glucose metabolism in



https://www.sciencealert.com/artificial-sweeteners-like-aspartame-can-be-toxic-to-our-gut-microbes

https://www.sciencealert.com/artificial-sweeteners-like-aspartame-can-be-toxic-to-our-gut-microbes

http://nationalpost.com/news/0717-na-sugar

https://www.theglobeandmail.com/news/national/sugar-substitutes-linked-to-weight-gain-and-health-problems-study-says/article35704562/

https://www.theglobeandmail.com/news/national/sugar-substitutes-linked-to-weight-gain-and-health-problems-study-says/article35704562/

https://www.heartandstroke.ca/-/media/pdf-files/canada/2017-position-statements/sugar-ps-eng.ashx

https://www.heartandstroke.ca/-/media/pdf-files/canada/2017-position-statements/sugar-ps-eng.ashx

https://health.gov/dietaryguidelines/2015/resources/DGA_Cut-Down-On-Added-Sugars.pdf

https://well.blogs.nytimes.com/2014/09/17/artificial-sweeteners-may-disrupt-bodys-blood-sugar-controls/

https://well.blogs.nytimes.com/2014/09/17/artificial-sweeteners-may-disrupt-bodys-blood-sugar-controls/

https://blogs.scientificamerican.com/guest-blog/artificial-sweeteners-may-have-despicable-impacts-on-gut-microbes/

https://blogs.scientificamerican.com/guest-blog/artificial-sweeteners-may-have-despicable-impacts-on-gut-microbes/

http://time.com/4859012/artificial-sweeteners-weight-loss/

http://news.nationalgeographic.com/news/2014/09/140917-sweeteners-artificial-blood-sugar-diabetes-health-ngfood/

http://news.nationalgeographic.com/news/2014/09/140917-sweeteners-artificial-blood-sugar-diabetes-health-ngfood/

https://www.thestar.com/life/2017/07/17/more-research-needed-into-the-effects-sugar-substitutes-have-on-health.html

https://www.thestar.com/life/2017/07/17/more-research-needed-into-the-effects-sugar-substitutes-have-on-health.html

http://www.webmd.com/food-recipes/news/20170717/do-artificial-sweeteners-raise-odds-for-obesity#1

http://www.webmd.com/food-recipes/news/20170717/do-artificial-sweeteners-raise-odds-for-obesity#1

https://www.researchnet-recherchenet.ca/rnr16/vwOpprtntyDtls.do?prog=2554&printfriendly=true

https://www.researchnet-recherchenet.ca/rnr16/vwOpprtntyDtls.do?prog=2554&printfriendly=true

142

healthy participants: the fructose and allulose catalytic effects (FACE) trial. Nutrients, 2018. 10(6): p. 750.

200. Health Canada, Draft guidance document on food health claims related to the reduction in post-prandial glycaemic response, Bureau of Nutritional Sciences, Editor. 2013, Health Products and Food Branch Editor.

201. Food and Drug Administration, Multiple endpoints in clinical trials: guidance for industry [draft guidance]. 2017.

202. Benjamini, Y.a.D.Y., The control of the false discovery rate in multiple testing under dependency. Annals of Statistics, 2001. 29(4): p. 1165-1188.

203. Benjamini, Y.a.Y.H., Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the royal statistical society, 1995: p. 289-300.

204. Leung, A.A., et al., Hypertension Canada's 2017 Guidelines for Diagnosis, Risk Assessment, Prevention, and Treatment of Hypertension in Adults. Can J Cardiol, 2017. 33(5): p. 557-576.

205. Matsuda, M.a.R.A.D., Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care, 1999. 22(9): p. 1462-1470.

206. Heart and Stroke Foundation of Canada. Healthy weight and waist. 2018; Available from: https://www.heartandstroke.ca/get-healthy/healthy-weight/healthy-weight-and-waist.

207. Nerenberg, K.A., et al., Hypertension Canada’s 2018 guidelines for diagnosis, risk assessment, prevention, and treatment of hypertension in adults and children. Canadian Journal of Cardiology, 2018. 34(5): p. 506-525.

208. Mouzaki, M., et al., Intestinal microbiota in patients with nonalcoholic fatty liver disease. Hepatology, 2013. 58(1): p. 120-7.

209. Caporaso, J.G., et al., Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. ISME J, 2012. 6(8): p. 1621-4.

210. Langille, M.G., et al., Predictive functional profiling of microbial communities using 16S rRNA marker gene sequences. Nat Biotechnol, 2013. 31(9): p. 814-21.

211. Kadish, A.H.a.H., D, A new method for the continuous monitoring of blood glucose by measurement of dissolved oxygen. Clinical Chemistry, 1965. 11(9): p. 869-875.

212. Mercodia, Mercodia insulin ELISA. 2008-2017: Uppsala, Sweden. 213. Wolever, T.M., et al., The glycemic index: methodology and clinical implications. Am J Clin Nutr,

1991. 54(5): p. 846-54. 214. Phillips, D.I., et al., Understanding oral glucose tolerance: comparison of glucose or insulin

measurements during the oral glucose tolerance test with specific measurements of insulin resistance and insulin secretion. Diabet Med, 1994. 11(3): p. 286-92.

215. Theytaz, F., et al., Metabolic fate of fructose ingested with and without glucose in a mixed meal. Nutrients, 2014. 6(7): p. 2632-49.

216. Logue, C., et al., Application of liquid chromatography–tandem mass spectrometry to determine urinary concentrations of five commonly used low-Calorie sweeteners: A novel biomarker approach for assessing recent intakes? Journal of Agricultural and Food Chemistry, 2017.

217. Schulz, K.F., D.G. Altman, and D. Moher, CONSORT 2010 Statement: updated guidelines for reporting parallel group randomised trials. BMJ, 2010. 340(mar23 1): p. c332-c332.

218. Government of Canada. Sugars. 2018; Available from: https://www.canada.ca/en/health-canada/services/nutrients/sugars.html.

219. Leung, A., et al., Hypertension Canada's 2017 guidelines for diagnosis, risk assessment, prevention, and treatment of hypertension in adults. Canadian Journal of Cardiology, 2017.

https://www.heartandstroke.ca/get-healthy/healthy-weight/healthy-weight-and-waist

https://www.canada.ca/en/health-canada/services/nutrients/sugars.html

https://www.canada.ca/en/health-canada/services/nutrients/sugars.html

Date post:	02-Oct-2021
Category:	Documents
Upload:	others
View:	2 times
Download:	0 times

Non-Nutritive Sweetened Beverages and Cardiometabolic Risk

Documents