Transcript UITNOD To open A dou in scho of suga bever D Op d 13: 4 Unive

A double-blind randomized trial in school children
on the effects of sugar-sweetened or sugar-free
beverages on body weight and body fatness
Janne de Ruyter
A double-blind randomized trial in school children
on the effects of sugar-sweetened or sugar-free
beverages on body weight and body fatness
The studies presented in this thesis were conducted at the Department of Health
Sciences and the EMGO+ Institute for Health and Care Research of the VU University
Amsterdam. The research described in this thesis was supported by a grant of the
Dutch Heart Foundation (DHF-2008B096), and the Netherlands Organization for
Health Research and Development (ZonMw-120520010). The DRINK study was also
supported by an Academy Professorship awarded by the Royal Netherlands Academy
of Arts and Sciences to prof.dr. M.B. Katan.
Financial support by the Dutch Heart Foundation for the publication of this thesis is
gratefully acknowledged.
Additional financial support for the publication of this thesis has been kindly provided
by the VU University.
Nederlandse titel: Een dubbelblinde gerandomiseerde studie naar het effect van
suikerhoudende of suikervrije dranken op lichaamsgewicht en vetmassa bij kinderen
Cover & layout : proefschrift-opmaken.nl
Printed by
: GVO drukkers & vormgevers B.V. | Ponsen & Looijen
ISBN
: 978 90 8659 654 6
© 2013, J. de Ruyter, Amsterdam, The Netherlands
All rights reserved. No part of this thesis may be reproduced, stored in a retrieval
system, or transmitted in any form or by any means, electronically, mechanically, by
photocopy, by recording, or otherwise, without prior permission of the author, or,
when appropriate, the publishers of the papers.
VRIJE UNIVERSITEIT
A double-blind randomized trial in school children on the
effects of sugar-sweetened or sugar-free beverages on
body weight and body fatness
ACADEMISCH PROEFSCHRIFT
ter verkrijging van de graad Doctor aan
de Vrije Universiteit Amsterdam,
op gezag van de rector magnificus
prof.dr. F.A. van der Duyn Schouten,
in het openbaar te verdedigen
ten overstaan van de promotiecommissie
van de Faculteit der Aard- en Levenswetenschappen
op donderdag 3 oktober 2013 om 13.45 uur
in de aula van de universiteit,
De Boelelaan 1105
door
Janne Catharine de Ruyter
geboren te Utrecht
promotor
: prof.dr. M.B. Katan
copromotor
: dr.ir. M.R. Olthof
Abstract
Background
The increased number of children who are overweight or obese is a major health problem, and
coincides with an increase in the consumption of sugar-sweetened beverages. Liquid calories
are thought to be more fattening than solid calories because they do not lead to a sense of
satiety. Therefore children may not reduce their intake of calories from other foods with a
resultant increase in total energy intake and weight gain. We designed the Double-blind,
Randomized Intervention study in Kids (DRINK) to examine the effect of masked replacement
of sugar-containing by sugar-free beverages on body weight, fat accumulation, satiety, liking
and wanting.
Methods
As preparation for DRINK, 89 children tasted 7 sugar-free (artificially-sweetened) and 7
matched sugar-sweetened beverages. The children liked both types equally well but were able
to discriminate between the two in 49% of the tests. However, 33% of the 49% correct
responses were made by chance alone. We therefore considered the 14 beverages eligible for
DRINK. However, we selected 3 pairs out of the 7 tested pairs for practical reasons. DRINK was
an 18-month trial involving 641 mostly normal-weight children from 4 to 11 years who were
habitual consumers of sugar-sweetened beverages. Children were randomly assigned to
receive 250 ml per day of a sugar-free, artificially-sweetened beverage (sugar-free group) or a
similar sugar-sweetened beverage that provided 104 kcal (sugar group). Beverages were
distributed through schools. We performed body, and sensory measurements at 0, 6, 12 and
18 months. At 18 months, 26% of the children had stopped consuming the beverages.
Results
Replacing sugar-containing beverages with sugar-free beverages reduced weight gain and fat
accumulation. The BMI z score increased on average by 0.02 SD units in the sugar-free group,
and by 0.15 SD units in the sugar group, mean difference -0.13 (95% confidence interval (CI) for
the difference, -0.21 to -0.05). Weight increased by 6.35 kg in the sugar-free group as
compared with 7.37 kg in the sugar group, mean difference -1.01 (95% CI for the difference, 1.54 to -0.48). Also, skinfold thicknesses (mm), waist-to-height ratio and fatmass (kg) according
to impedance increased significantly less in the sugar-free group than in the sugar group, −2.2
(95% CI −4.0 to −0.4), −0.4 (95% CI −1.0 to −0.0), and −0.57 (95% CI −1.02 to −0.12)
respectively. The difference in satiety between both groups was not statistically significant; the
adjusted odds ratio for a 1 point increase on a 5-point scale in satiety in the sugar group versus
the sugar-free group was 0.77 at 1 minute (95% CI, 0.46 to 1.29), and 1.44 at 15 minutes after
intake (95% CI, 0.86 to 2.40). The sugar-group liked and wanted their beverage slightly more
than the sugar-free group, adjusted odds ratio 1.63 (95% CI 1.05 to 2.54) and 1.65 (95% CI 1.07
to 2.55), respectively.
Conclusion
We showed that masked replacement of sugar-containing beverages with sugar-free beverages
reduced weight gain and fat accumulation, and produced similar degrees of satiety. A plausible
explanation is that the removal of liquid sugar was indeed not sensed by satiating mechanisms
and was therefore incompletely compensated for by the increased consumption of other
foods. Sugar-free beverages may therefore contribute to reducing weight gain in children.
Contents
CHAPTER 1
General introduction
11
CHAPTER 2
A sensory taste study of sugar-sweetened versus artificiallysweetened beverages among children aged between 5 and
12 years
25
CHAPTER 3
Effect of sugar-sweetened beverages on body weight in
children: design and baseline characteristics of the Doubleblind, Randomized Intervention study in Kids
43
CHAPTER 4
A trial of sugar-free or sugar-sweetened drinks and body
weight in children
71
CHAPTER 5
The effect of sugar-free versus sugar-sweetened beverages
on satiety, liking and wanting: an 18 month randomized
double-blind trial in children
105
CHAPTER 6
General discussion
135
SUMMARY
155
SAMENVATTING
161
DANKWOORD
166
ABOUT THE AUTHOR
168
PUBLICATIONS & PRESENTATONS
169
CHAPTER 1
General introduction
CHAPTER 1
This thesis is built around the DRINK trial: The Double blind Randomized Intervention
study in Kids (DRINK). The aim was to answer the question “Does the consumption of
sugary drinks cause weight gain in children?”
Overweight and obesity
The prevalence of obesity and overweight in children has become a major public
health problem worldwide. Recently, the prevalence of high BMI (Body Mass Index,
the weight in kilograms divided by the square of the height in meters) in children
appeared to plateau,1 although the number of overweight children remains high. In
2009-2010, the percentage of children with overweight or obesity was 31.8% in the
USA.2 Overweight was defined here as at or above the sex-specific 85th percentile on
the CDC's 2000 BMI-for-age growth charts but less than the 95th percentile; obesity
was defined as a BMI at or above the sex-specific 95th percentile.3 This means that
the prevalence of overweight and obesity did not change compared with 2007-2008.4
In 2009, 12.8% of Dutch boys and 14.8% of Dutch girls aged 2-21 years were
overweight and 1.8% of boys and 2.2% of girls were classified as truly obese. This
reflects a two to three fold higher prevalence of overweight and four to six fold
increase of obesity in the Netherlands since 1980.5
Overweight and obesity in childhood are associated with health risks in
childhood and also later in life. Obesity in childhood is a risk factor for adult obesity,
type 2 diabetes, cancer, cardiovascular diseases and death before of the age of 55
years.6-11 Obesity in children also has negative health consequences during childhood
itself such as insulin resistance, hypertension, dyslipidemia and type 2 diabetes.12 In
addition, children who are overweight are often stigmatized and bullied by other
children.13 Thus, prevention of overweight and obesity at a young age is crucial to the
prevention of chronic diseases both in childhood and in adulthood.
A simple explanation of obesity is that it is caused by an imbalance between
energy intake and energy expenditure. However, the issue appears to be more
complex than this, because the imbalance between energy intake and expenditure
has been associated with many factors. Swinburn et al. proposed an ecological model
for understanding obesity (Fig. 1.1). This model incorporates environmental as well as
biological and behavioral influences.14 Biological factors include gender, ethnic origin,
age, pubertal stage, and physical fitness. Behavioral factors include socioeconomic
status, parenting, snacking, television viewing, lack of exercise, and peer group.
Environmental factors are factors such as the education and health system, and the
food industry.
12
GENERAL INTRODUCTION
Figure 1.1 An ecological model for understanding obesity (PA, physical activity). 14
‘Obesogenic environment’
Why obesity rates have risen so rapidly is widely debated. Perhaps it is our
environment that is to blame; obesity may represent ‘a normal reaction to an
abnormal environment’.15 The environment of many developed countries may be
described as an ‘obesogenic environment’. This can be defined as the ‘the sum of
influences that the surroundings, opportunities, or conditions of life have on
promoting obesity in individuals or populations’.14 An obesogenic environment
promotes a high energy intake and sedentary behaviors. The challenge is to create an
environment where the healthy choice is the easy choice. Examples include creating
new parts of town where shops are at walking or cycling distances and discouraging
excessive space for parking cars, and the provision of safe walkways to school for
children.
Research in adults has shown that small changes at population level may stop
the obesity epidemic.16 This implies that it makes sense to regard obesity as a public
health problem rather than solely an individual matter.17 Focusing on individual
lifestyle factors without addressing environmental factors is insufficient to reverse the
obesity epidemic.18 Interventions that target the obesogenic environment may help
people to improve their dietary habits and levels of physical activity.19 Hypothetically,
if sugar-sweetened beverages contribute to the increase in overweight or obesity,
interventions should focus on the removal of such beverages and the promotion of
the consumption of sugar-free beverages in children’s environment.
Consumption of beverages
There is a lot of speculation about which foods are particularly fattening; sugarsweetened beverages are seen as major culprits.20 In the last decade, the
13
CHAPTER 1
consumption of sugar-sweetened beverages has increased.21 Sugar intake from sugarsweetened beverages alone, the largest single caloric food source in the USA,
approaches 15% of the daily caloric intake for some Americans.22,23 Currently, 81% of
USA school children consume at least one sugar-sweetened beverage per day, and the
daily average intake is 272 fluid kcal.22 Dutch children drink approximately 0.7 liter
sugary beverages per day which equals approximately 280 calories.24 Children aged 9
to 13 years old need approximately 2100 kcal per day.25 This means that liquid
calories account for more than 10% of their daily intake. Intake of sugar-sweetened
drinks does not differ in children with high or low social economic background.26
Sugar-sweetened beverages cover a large range of products, e.g. still and carbonated
soft drinks, fruit juice-based drinks, ‘energy drinks’, pure fruit juices, fruit drinks,
‘vitamin water drinks’, lemonades, iced and sweetened tea, and flavored milk drinks.
To add to the complexity, the consumption of sugar-free beverages has also been
associated with overweight. It has been suggested that non-caloric sweeteners
increase the sense of hunger and ensuing energy intake.27 This theory was initially
put forward by Blundell and Hill,28 but refuted by most other research.29-31
Some theories suggest that liquid calories are considered more fattening than
solid foods because they do not lead to a sense of satiety.32 This would imply that
liquid calories are not ‘sensed’, and that liquid calories do not satiate.33 Consequently,
sugar-sweetened and sugar-free beverages would produce similar satiety. Therefore,
when children are given sugar-free instead of sugar-containing drinks they may not
make up the missing calories from other sources with a resultant decrease in total
energy intake and weight gain. However, solid evidence for the liquid sugar theory is
lacking.
Sugar-sweetened beverages and overweight
The belief that liquid calories are particularly fattening has already been translated
into government nutrition policy despite that fact that evidence from properly
conducted clinical trials was almost non-existent. Consumption of sugar-sweetened
beverages has been associated with weight gain in most20,34-37 though not all
observational studies.38,39 However, children who drink more sugar-sweetened
beverages also tend to eat more fast food and to watch more television.40 Most
studies adjust statistically for such confounders, but residual and unmeasured
confounding cannot be ruled out.41,42 Some experiments have addressed the effect of
sugar-free versus sugar-sweetened beverages on body weight, food intake or hunger
(Table 1.1). The results on both food intake and hunger,43-48 and body weight49-51 were
14
GENERAL INTRODUCTION
Table 1.1 Experimental studies on the effect of sugar-free beverages versus sugarsweetened beverages on food intake, hunger, or body weight. Children or adults
were given sugar-free beverages or sugar-sweetened beverages for a long or short
time. Researchers then measured how much the participants ate after intake, their
feeling of hunger or body weight.
Effect of sugar-free beverages versus sugar-sweetened beverages on food intake and/or
hunger
First author
46
Rodin, 1990
47
Rolls et al., 1990
Canty et al., 1991
44
Beridot-Therond
43
et al., 1998
45
Holt et al., 2000
Van Wymelbeke
48
et al., 2004
Outcome
Effect
Food intake
Food intake
Appetite
Food intake
Hunger
Food intake
Same intake
Same intake
Same degree of appetite
Same intake
More hunger on sugar-free drinks
Same intake
Food intake
Food intake
Hunger
Same intake
Same intake
Same degree of hunger
Effect of sugar-free beverages versus sugar-sweetened beverages on body weight
49
Tordoff et al., 1990
53
James et al., 2004
Ebbeling et al., 2006
Sichieri et al., 2009
50
Tate et al., 2012
a
51
54
Body weight
Number of overweight
children
a
BMI
BMI
Body weight
Weight loss on sugar-free drinks
Sugar-free drinks decreased % of
children with obesity and overweight
Weight loss on sugar-free drinks only
in those in upper baseline BMI tertile
Weight loss on sugar-free drinks only
in girls
Sugar-free drinks increased likelihood
to achieve a 5% weight loss
Body-Mass Index, the weight in kilograms divided by the square of the height in meters
inconclusive, either by their design, sample size, lack of adequate placebos, duration,
lack of appropriate randomization, or a combination of these factors. In addition,
there have been two trials in which whole school classes were either encouraged to
drink less soda or received no special instruction.53,54 Such studies are not equivalent
to a trial where each child in a class may receive something different. Therefore they
were considered unconvincing52 and produced equivocal results. James et al.53 found
that a school-based educational program aimed at reducing consumption of
15
CHAPTER 1
carbonated drinks did not affect BMI, although the authors did claim that it reduced
the proportion of overweight and obese children. Also the study of Sichieri et al.54
discouraged students from drinking sugar-sweetened beverages with an educational
program. They found that decreasing sugar-sweetened beverages intake significantly
reduced BMI only among girls. All in all, conclusive evidence was lacking to show that
the replacement of sugar-containing beverages with noncaloric beverages would
diminish weight gain.
The DRINK trial
We designed the DRINK trial: The double blind randomized intervention study in kids
(DRINK) in accordance with the guidelines for evidence based medicine.55 DRINK
investigated the effect of masked replacement of sugar-sweetened with sugar-free
drinks on body weight. Proper blinding was crucial because food choice and caloric
intake are sensitive to social cues and expectations.56 Also cognition and feelings have
a considerable impact on such outcomes. Children may eat less not because of the
sugar content of the beverage but because they receive explicit or subconscious cues
that they are expected to lose weight. The specific effect of liquid sugar content on
body weight can be determined only when all drinks look and taste the same.56 The
trial was therefore a strict test of the effects of liquid calories on body weight through
unconscious physiological mechanisms that regulate food intake, independent of
behavioral, cognitive, and psychological factors.
The development and manufacturing of sugar-free and sugar-sweetened
beverages, that tasted and looked the same, was a major challenge in our study. The
beverages needed to be safe and stable, attractive and pleasant-tasting to children.
We manufactured sugar-sweetened beverages containing 26 g of sugar and 104 kcal
per 250-mL can. The sugar-free beverages contained the artificial sweeteners
sucralose (0.157 g/L) and acesulfame potassium (0.057 g/L). Sensory studies had
shown that sucralose does not have the bitter aftertaste attributed to some other
non-nutritive sweeteners.57,58 Sucralose and acesulfame potassium are also highly
stable.59 These sweeteners have been approved by the US Food and Drug
Administration, the Joint Commission of Experts on Food Additives of the World
Health Organization, the Food and Agriculture Organization, and the European Food
Safety Authority. The acceptable daily intake (ADI) for sucralose is 15 mg/kg
bodyweight/day.60 This means that a child weighing 20 kg can drink 7 cans of our
sugar-free study beverage per day. The ADI for acesulfame potassium is 9 mg/kg
16
GENERAL INTRODUCTION
bodyweight/day which corresponds to 13 cans.61 We developed and manufactured
beverages in four flavors: lemon, mango, peach and raspberry.
Outline of this thesis
This thesis aims to examine the long term effects of masked replacement of sugarsweetened beverages with sugar-free beverages on body weight and fat accumulation
in children. The secondary aim was to examine the long-term effects of such
beverages on the degree of satiety, liking and wanting. In the preparation of DRINK,
we performed three pilot studies. One pilot study was done to select appropriate
beverages for DRINK. The results of this pilot study will be described in Chapter 2. The
other two pilot studies examined logistical issues and the accurate time point to
collect urine samples as compliance markers. The results of these will not be
described in this thesis. Chapter 3 elaborates on the design of DRINK, and chapter 4
describes the effects of the sugar-free and the sugar-sweetened beverages on body
fat and fat accumulation. The effects of such beverages on satiety, liking and wanting
are described in chapter 5. Chapter 6 reflects on the results described in this thesis.
17
CHAPTER 1
References
1. Olds T, Maher C, Zumin S, et al. Evidence that the prevalence of childhood
overweight is plateauing: data from nine countries. Int J Pediatr Obes 2011;6:342-60.
2. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of obesity and trends in
body mass index among US children and adolescents, 1999-2010. JAMA
2012;307:483-90.
3. Ogden CL, Flegal KM. Changes in terminology for childhood overweight and
obesity. Natl Health Stat Report 2010:1-5.
4. Ogden CL, Carroll MD, Curtin LR, Lamb MM, Flegal KM. Prevalence of high body
mass index in US children and adolescents, 2007-2008. JAMA 2010;303:242-9.
5. Schonbeck Y, Talma H, van Dommelen P, et al. Increase in prevalence of
overweight in dutch children and adolescents: a comparison of nationwide growth
studies in 1980, 1997 and 2009. PLoS One 2011;6:e27608.
6. Steinberger J, Daniels SR. Obesity, insulin resistance, diabetes, and
cardiovascular risk in children: an American Heart Association scientific statement
from the Atherosclerosis, Hypertension, and Obesity in the Young Committee (Council
on Cardiovascular Disease in the Young) and the Diabetes Committee (Council on
Nutrition, Physical Activity, and Metabolism). Circulation 2003;107:1448-53.
7. Magarey AM, Daniels LA, Boulton TJ, Cockington RA. Predicting obesity in early
adulthood from childhood and parental obesity. Int J Obes Relat Metab Disord
2003;27:505-13.
8. Lobstein T, Baur L, Uauy R. Obesity in children and young people: a crisis in
public health. Obes Rev 2004;5 Suppl 1:4-82.
9. Hayman LL, Meininger JC, Daniels SR, et al. Primary prevention of cardiovascular
disease in nursing practice: focus on children and youth: a scientific statement from
the American Heart Association Committee on Atherosclerosis, Hypertension, and
Obesity in Youth of the Council on Cardiovascular Disease in the Young, Council on
Cardiovascular Nursing, Council on Epidemiology and Prevention, and Council on
Nutrition, Physical Activity, and Metabolism. Circulation 2007;116:344-57.
10. Hannon TS, Rao G, Arslanian SA. Childhood obesity and type 2 diabetes mellitus.
Pediatrics 2005;116:473-80.
11. Franks PW, Hanson RL, Knowler WC, Sievers ML, Bennett PH, Looker HC.
Childhood obesity, other cardiovascular risk factors, and premature death. N Engl J
Med 2010;362:485-93.
18
GENERAL INTRODUCTION
12. Weiss R, Dziura J, Burgert TS, et al. Obesity and the metabolic syndrome in
children and adolescents. N Engl J Med 2004;350:2362-74.
13. Janssen I, Craig WM, Boyce WF, Pickett W. Associations between overweight and
obesity with bullying behaviors in school-aged children. Pediatrics 2004;113:1187-94.
14. Swinburn B, Egger G, Raza F. Dissecting obesogenic environments: the
development and application of a framework for identifying and prioritizing
environmental interventions for obesity. Prev Med 1999;29:563-70.
15. Rutter H. Where next for obesity? Lancet 2011;378:746-7.
16. Hill JO. Can a small-changes approach help address the obesity epidemic? A
report of the Joint Task Force of the American Society for Nutrition, Institute of Food
Technologists, and International Food Information Council. Am J Clin Nutr
2009;89:477-84.
17. Veerman JL, Barendregt JJ, van Beeck EF, Seidell JC, Mackenbach JP. Stemming
the obesity epidemic: a tantalizing prospect. Obesity (Silver Spring) 2007;15:2365-70.
18. Hill JO, Wyatt HR, Reed GW, Peters JC. Obesity and the environment: where do
we go from here? Science 2003;299:853-5.
19. Hill JO, Peters JC. Environmental contributions to the obesity epidemic. Science
1998;280:1371-4.
20. Malik VS, Schulze MB, Hu FB. Intake of sugar-sweetened beverages and weight
gain: a systematic review. Am J Clin Nutr 2006;84:274-88.
21. Duffey KJ, Popkin BM. Shifts in patterns and consumption of beverages between
1965 and 2002. Obesity (Silver Spring) 2007;15:2739-47.
22. Wang YC, Bleich SN, Gortmaker SL. Increasing caloric contribution from sugarsweetened beverages and 100% fruit juices among US children and adolescents, 19882004. Pediatrics 2008;121:e1604-14.
23. Block G, Mandel R, Gold E. On food frequency questionnaires: the contribution of
open-ended questions and questions on ethnic foods. Epidemiology 2004;15:216-21.
24. Brug J, van Stralen MM, Te Velde SJ, et al. Differences in weight status and
energy-balance related behaviors among schoolchildren across Europe: the ENERGYproject. PLoS One 2012;7:e34742.
25. Voedingscentrum (English translation: Netherlands Nutrition Center). Hoe eet ik
niet te veel? (English translation: How do I not eat too much).
http://www.voedingscentrum.nl/. Accessed March 13, 2013.
26. Buurma-Rethans E, Fransen H, Van Rossum C. Consumptie van dranken met
toegevoegd suiker door kinderen van de basisschoolleeftijd. Resultaten van VCP 2007-
19
CHAPTER 1
2010 (English translation: Consumption of beverages with added sugars in school
children. Dutch National Food Consumption Survey Á young children 2007-2010)
27. Rogers PJ, Carlyle JA, Hill AJ, Blundell JE. Uncoupling sweet taste and calories:
comparison of the effects of glucose and three intense sweeteners on hunger and
food intake. Physiol Behav 1988;43:547-52.
28. Blundell JE, Hill AJ. Paradoxical effects of an intense sweetener (aspartame) on
appetite. Lancet 1986;1:1092-3.
29. Bellisle F, Drewnowski A. Intense sweeteners, energy intake and the control of
body weight. Eur J Clin Nutr 2007;61:691-700.
30. Mattes RD, Shikany JM, Kaiser KA, Allison DB. Nutritively sweetened beverage
consumption and body weight: a systematic review and meta-analysis of randomized
experiments. Obes Rev 2011;12:346-65.
31. Rolls BJ. Effects of intense sweeteners on hunger, food intake, and body weight:
a review. Am J Clin Nutr 1991;53:872-8.
32. Mattes R. Fluid calories and energy balance: the good, the bad, and the
uncertain. Physiol Behav 2006;89:66-70.
33. Bellisle F, Drewnowski A, Anderson GH, Westerterp-Plantenga M, Martin CK.
Sweetness, satiation, and satiety. J Nutr 2012;142:1149S-54S.
34. Ludwig DS, Peterson KE, Gortmaker SL. Relation between consumption of sugarsweetened drinks and childhood obesity: a prospective, observational analysis. Lancet
2001;357:505-8.
35. Dubois L, Farmer A, Girard M, Peterson K. Regular sugar-sweetened beverage
consumption between meals increases risk of overweight among preschool-aged
children. J Am Diet Assoc 2007;107:924-34.
36. Dhingra R, Sullivan L, Jacques PF, et al. Soft drink consumption and risk of
developing cardiometabolic risk factors and the metabolic syndrome in middle-aged
adults in the community. Circulation 2007;116:480-8.
37. Schulze MB, Manson JE, Ludwig DS, et al. Sugar-sweetened beverages, weight
gain, and incidence of type 2 diabetes in young and middle-aged women. JAMA
2004;292:927-34.
38. O'Connor TM, Yang SJ, Nicklas TA. Beverage intake among preschool children
and its effect on weight status. Pediatrics 2006;118:e1010-8.
39. Newby PK, Peterson KE, Berkey CS, Leppert J, Willett WC, Colditz GA. Beverage
consumption is not associated with changes in weight and body mass index among
low-income preschool children in North Dakota. J Am Diet Assoc 2004;104:1086-94.
20
GENERAL INTRODUCTION
40. Park S, Blanck HM, Sherry B, Brener N, O'Toole T. Factors associated with sugarsweetened beverage intake among United States high school students. J Nutr
2012;142:306-12.
41. Vandenbroucke JP. Observational research, randomised trials, and two views of
medical science. PLoS Med 2008;5:339-43.
42. Fewell Z, Davey Smith G, Sterne JAC. The Impact of Residual and Unmeasured
Confounding in Epidemiologic Studies: A Simulation Study. Am J Epidemiol
2007;166:646-55.
43. Beridot-Therond ME, Arts I, Fantino M, De La Gueronniere V. Short-term effects
of the flavour of drinks on ingestive behaviours in man. Appetite 1998;31:67-81.
44. Canty DJ, Chan MM. Effects of consumption of caloric vs noncaloric sweet drinks
on indices of hunger and food consumption in normal adults. Am J Clin Nutr
1991;53:1159-64.
45. Holt SH, Sandona N, Brand-Miller JC. The effects of sugar-free vs sugar-rich
beverages on feelings of fullness and subsequent food intake. Int J Food Sci Nutr
2000;51:59-71.
46. Rodin J. Comparative effects of fructose, aspartame, glucose, and water preloads
on calorie and macronutrient intake. Am J Clin Nutr 1990;51:428-35.
47. Rolls BJ, Kim S, Fedoroff IC. Effects of drinks sweetened with sucrose or
aspartame on hunger, thirst and food intake in men. Physiol Behav 1990;48:19-26.
48. Van Wymelbeke V, Beridot-Therond ME, de La Gueronniere V, Fantino M.
Influence of repeated consumption of beverages containing sucrose or intense
sweeteners on food intake. Eur J Clin Nutr 2004;58:154-61.
49. Tordoff MG, Alleva AM. Effect of drinking soda sweetened with aspartame or
high-fructose corn syrup on food intake and body weight. Am J Clin Nutr 1990;51:963-9.
50. Tate DF, Turner-McGrievy G, Lyons E, 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;95:555-63.
51. Ebbeling CB, Feldman HA, Osganian SK, Chomitz VR, Ellenbogen SJ, Ludwig DS.
Effects of decreasing sugar-sweetened beverage consumption on body weight in
adolescents: a randomized, controlled pilot study. Pediatrics 2006;117:673-80.
52. French SA, Hannan PJ, Story M. School soft drink intervention study. BMJ
2004;329:E315-6.
21
CHAPTER 1
53. James J, Thomas P, Cavan D, Kerr D. Preventing childhood obesity by reducing
consumption of carbonated drinks: cluster randomised controlled trial. BMJ
2004;328:1237-43.
54. Sichieri R, Paula Trotte A, de Souza RA, Veiga GV. School randomised trial on
prevention of excessive weight gain by discouraging students from drinking sodas.
Public Health Nutr 2009;12:197-202.
55. Sackett DL, Rosenberg WM, Gray JA, Haynes RB, Richardson WS. Evidence based
medicine: what it is and what it isn't. BMJ 1996;312:71-2.
56. Katan MB. Weight-loss diets for the prevention and treatment of obesity. N Engl
J Med 2009;360:923-5.
57. Wiet SG, Beyts PK. Sensory Characteristics of Sucralose and Other High-Intensity
Sweeteners. J Food Sci 1992;57:1014-9.
58. Kuhn C, Bufe B, Winnig M, et al. Bitter taste receptors for saccharin and
acesulfame K. J Neurosci 2004;24:10260-5.
59. Malik A, Jeyarani T, Raghavan B. A comparison of artificial sweeteners' stability
in a lime-lemon flavored carbonated beverage. Journal of Food Quality 2002;25:75-82.
60. Scietific Committee on Food. Sucralose (opinion adapted by the SCF on 7
September 2000). http://ec.europa.eu/food/fs/sc/scf/out68_en.pdf. Accessed at
February 19, 2011.
61. Scietific Committee on Food. Re-evaluation of acesulfame K with reference to
the previous SCF opinion of 1991 (opinion expressed on 9th March 2000).
http://ec.europa.eu/food/fs/sc/scf/out52_en.pdf. Accessed at February 19, 2011.
22
CHAPTER 2
A sensory taste study of sugar-sweetened versus
artificially-sweetened beverages among children aged
between 5 and 12 years
Janne C de Ruyter
Rosa Kas
Margreet R Olthof
Martijn B Katan
Submission in preparation
CHAPTER 2
Abstract
Background
Previous trials have shown that masked replacement of sugar-sweetened by sugarfree beverages reduced weight gain in children. This suggests that children may
benefit from drinking more sugar-free beverages, such as water or artificiallysweetened beverages, to avoid weight gain. However, this would require children to
like artificially-sweetened beverages as much as they like sugar-sweetened beverages.
Studies that have examined the preference in children for artificially-sweetened
beverages over sugar-sweetened beverages are scarce. The aim of this study was to
examine children’s preference towards artificially-sweetened beverages and matching
sugar-sweetened beverages, and children’s ability to discriminate between the two
versions.
Methods
89 children aged between 5 and 12 tasted seven artificially-sweetened beverages and
seven matched sugar-sweetened beverages, for a total of 14 beverages. We used a 5point scale to measure how much the children liked each beverage, and Triangle tests
to check their ability to discriminate between the matched versions.
Results
The mean (±SD) score for how much the children liked the artificially-sweetened
beverages, on a 5-point scale, was 3.39±0.7, and the score for the sugar-sweetened
beverages was 3.39±0.6. The difference in the scores was not statistically significant
(P=0.9). Children were able to discriminate between sugar-sweetened and artificiallysweetened beverages: more children identified the ‘different’ samples than would
have been expected by chance alone. In total, the 89 children performed 619 Triangle
tests. In 306 (49%) of these tests they identified the ‘different’ sample; this is 100
more than the 206 correct responses expected by chance alone (P = 0.005).
Conclusions
We found that children liked artificially-sweetened and sugar-sweetened beverages
equally, but were able to discriminate between the two. Recently, we have shown
that sugar-free beverages reduce weight compared with sugar-sweetened beverages.
Although water is the preferred option, artificially-sweetened beverages provide a
tasty alternative.
26
SENSORY TASTE STUDY
Introduction
The increasing prevalence of obesity in children is a major health problem,1,2 and
coincides with an increase in the consumption of sugar-sweetened beverages.3 Recent
trials have shown that replacement of sugar-sweetened beverages by sugar-free
beverages reduces weight gain in children.4,5 Although water is the preferred sugarfree option, artificially-sweetened beverages may provide a healthy alternative to
sugar-sweetened beverages. In recent decades, children’s consumption of artificiallysweetened beverages has increased from less than 1% of the total beverages
consumed in 1999 to 2000 to more than 7% in 2007 to 2008.6 In order to persuade
children to drink an artificially-sweetened beverage instead of the sugar variant, they
need to like the artificially-sweetened beverage as much as the sugar-sweetened
beverage they are used to.
Little is known about the taste for artificially-sweetened beverages in children.
Results from trials that measured whether adults like them are inconsistent. A threeweek study in adults comparing beverages sweetened by aspartame with those
sweetened by sugar found that ratings of pleasantness were similar for both types.7 In
contrast, a four-week study in adults found that beverages sweetened with a blend of
aspartame, acesulfame potassium, and saccharin were rated lower in pleasantness
than beverages containing sucrose.8 No study on preferences for artificiallysweetened and sugar-sweetened beverages in children has yet been carried out.
Also, little is known about the ability of children to discriminate between
artificially-sweetened beverages and sugar-sweetened beverages. The food industry
invests in producing artificially-sweetened beverages that are indistinguishable from
sugar-sweetened beverages. This is challenging because an acceptable ‘mouth feel’ is
related to sugar content and is difficult to mimic with artificial sweeteners.9 The taste
of most artificial sweeteners is different from the taste of sucrose, possibly due to
bitterness10 and non-sweet aftertastes.11-13 Previous studies in adults suggest that the
beverages with the most similar taste to beverages sweetened with sucrose are
beverages sweetened with aspartame, followed by beverages sweetened by
cyclamate, saccharin and acesulfame potassium.14-18 Blends of artificial sweeteners
tend to be less similar to sucrose than aspartame, but they increase the similarity
between the artificial sweeteners and sucrose.15,19,20 No study examining the ability of
children to discriminate between sugar-sweetened and artificially-sweetened
beverages has yet been performed.
27
CHAPTER 2
The aim of the present study was to examine how much children liked artificiallysweetened beverages and matching sugar-sweetened beverages, and their ability to
discriminate between them.
Methods
Design
We performed a single-blind sensory study in November 2008 among 89 children
aged 5 to 12 years. The study protocol was approved by the Medical Ethical
Committee of the VU University Amsterdam.
Study population
We recruited children at an elementary school in the town of Purmerend, located in
an urbanized area near Amsterdam. Prior to this study, one researcher recorded what
the children drank at this school during meal breaks in three grades: one grade with
22 children aged 4 to 6 years, one with 25 children aged 6 to 8 years, and one with 27
children aged 9 to 11 years old. We found that 136 (92%) of the 148 beverages
consumed were sugar-sweetened beverages, and only 12 (8%) were water, milk or
artificially-sweetened beverages. We concluded that the children were habitual
consumers of sugar-sweetened beverages. We then sent out information letters
about this study to parents of all 262 children in the elementary school. This letter
also contained an informed consent form. Written informed consent was obtained
from a parent or guardian. A total of 89 (34%) children were willing and able to
participate, and were enrolled because there were no exclusion criteria.
Beverages
We used 14 beverages, seven pairs of artificially-sweetened and sugar-sweetened
beverages. Four pairs were available in the supermarkets, and three were noncommercially custom-made (Table 2.1). We selected these beverages because each
pair included an artificially-sweetened and a sugar-sweetened beverage of the same
flavor: H.J. Heinz Food Company (Pittsburg, U.S.A.) produced two pairs (Roosvicee
beverages, with forest fruits and peach flavors), Spadel Group (Brussels, Belgium)
produced two pairs (Spa beverages, with apple/cherry and forest fruits flavors), and
Unilever (Colworth, U.K.) produced three pairs of non-commercially custom-made
beverages, with lemon, mango, and peach flavors. The artificially-sweetened
beverages contained a blend of artificial sweeteners: sucralose plus acesulfame
28
SENSORY TASTE STUDY
potassium or cyclamate plus acesulfame potassium plus saccharin (Table 2.1). The
artificially-sweetened beverages contained 0.1% to 3.5% sugar (Table 2.1). The
beverages were non-carbonated, which facilitates the perception of taste.15
Table 2.1 Energy value, amount of sugar and artificial-sweeteners of the 14 beverages.
Type of
Brand name
Flavour
Energy
Sucrose Artificial sweeteners
beverage
(kcal/100 (g/100 (g/L)
ml)
ml)
Artificiallysweetened
Non-commercially
a
custom-made
Non-commercially
a
custom-made
Non-commercially
a
custom–made
b
Roosvicee
b
Roosvicee
Sugarsweetened
Lemon
1
0.1
Mango
1
0.1
Peach
1
0.1
Forest fruits
12
2.9
Peach
14
3.5
Spa
c
Apple/Cherry
8
1.7
Spa
c
Forest fruits
8
1.7
0.135 Sucralose, 0.050
acesulfame potassium
0.135 Sucralose, 0.050
acesulfame potassium
0.135 Sucralose, 0.050
acesulfame potassium
Sucralose acesulfame
d
potassium
Cyclamate, acesulfame
d
potassium, saccharin
Cyclamate, acesulfame
d
potassium, saccharin
Cyclamate, acesulfame
d
potassium, saccharin
0
Non-commercially Lemon
35
8
a
custom-made
Non-commercially Mango
35
8
0
a
custom-made
Non-commercially Peach
35
8
0
a
custom-made
b
Roosvicee
Forest fruits
39
9.6
0
b
Roosvicee
Peach
40
9.9
0
c
Spa
Apple/ Cherry 42
10.3
0
c
Spa
Forest fruits
36
8.8
0
a
Produced by Unilever (Colworth, U.K. )
bb
**
Produced by H.J. Heinz Food Company (Pittsburg, U.S.A.)
cc
** Produced by the Spadel Group (Brussels, Belgium)
dd
**
Exact amounts of artificial-sweeteners were not provided by the manufactures
29
CHAPTER 2
Procedure
We performed the tests during school hours in the staff room of the elementary
school. The tests lasted approximately 15 minutes for each child. We performed two
tests: the Pleasantness test and the Triangle test (described in detail below). A total of
45 children first performed the Pleasantness test and then the Triangle test; 44
children performed the tests the other way around. We used transparent 25 mL
medicine cups that we filled up to 15 mL. We offered the beverages at room
temperature. Children took a sip, tasted the beverage, and swallowed down the
liquid. Children were allowed to re-taste if necessary and could drink water in
between different beverages, but this was not mandatory. The Pleasantness test has
been validated for children in this age group,21 and the Triangle test has also been
used for earlier research in children.22
Sensory tests
Pleasantness test
We placed 16 cups on a tray for the Pleasantness test, 14 for the beverages that we
were testing plus two practice samples (1 x water, 1 x sugar-sweetened beverage).
We first offered the two practice samples, and then offered the 14 beverages in the
test, one-by-one in a predetermined order as explained below. Pleasantness was
rated using a 5-point scale questionnaire with five faces representing a range of liking.
We asked the children to point out the face that indicated how much they liked the
beverage (Table 2.2).
Triangle test
We placed 24 cups on a tray for the Triangle test, i.e. for each pair, three samples (7 x
3 = 21 samples) plus one set of three practice samples (2 x sugar-sweetened, 1 x
artificially-sweetened). Prior to the test, we showed the children three shapes, two
squares and one triangle, and asked children to indicate the ‘different’ shape. This
clarified the concept of ‘different’. We then offered the practice samples, followed by
the seven sets of the beverages in the test in a predetermined order as explained
below. For each set, we asked the children to taste the three samples and indicate the
‘different’ sample by pointing out the different sample.
Sequence of the beverages
For the Pleasantness test, we generated a unique sequence of the 14 beverages for
each child.23 For the Triangle test, all children received the seven beverages in the
same non-random order: Roosvicee forest fruits, Roosvicee peach, Non-commercially
30
SENSORY TASTE STUDY
custom-made lemon, Non-commercially custom-made peach, Non-commercially
custom-made mango, Spa forest fruits, and Spa apple/cherry. However, for each of
the seven beverages, the order of the three samples varied for each beverage and had
any of the six possible sequences BAA, ABA, AAB, ABB, BAB, BBA where A denotes the
sugar-sweetened version, and B denotes the artificially-sweetened version. One of the
six sequences was randomly allocated to each of the seven beverages.23
Statistical Analyses
For all analyses of the Pleasantness test, we performed Wilcoxon signed rank tests
because the data were not fully normally distributed. We combined the data for the
seven artificially-sweetened beverages and the seven sugar-sweetened beverages,
and examined whether the children liked both versions equally. We also examined
whether the children liked both versions within each pair of beverages equally. We
also combined the beverages with the same blends of artificial sweeteners, and
examined whether the children liked the sugar-sweetened and artificially-sweetened
versions equally.
For the Triangle test, we performed a one-sided binomial test (π = 1/3). We
combined the data for the seven artificially-sweetened and seven sugar-sweetened
beverages to examine whether the children discriminated between the artificiallysweetened and sugar-sweetened beverages. We also examined whether the children
discriminated between both versions within each pair of beverages. We also
combined the beverages with the same blends of artificial-sweeteners, and examined
whether the children discriminated between the two versions.
We used the Statistical Package for the Social Sciences (SPSS) version 17.0 to
perform the analyses. The level of significance was P<0.05 (Pleasantness test: 2-tailed
tests, Triangle test: 1-tailed test).
31
CHAPTER 2
Results
Participants
The age of the 89 participants was 9.2±2.0 years (mean±SD) at the start of the study.
We enrolled 51 (57%) girls. We did not collect other baseline characteristics.
However, we know that the school was located in an area with a mean socioeconomic status z score of 0.51.24 We calculated this z score with a standardized
Dutch database that includes the Dutch zip codes with their socio-economic status.
For example, an upper class neighborhood in Amsterdam has a socio-economic status
z score of 2.97 while a lower class neighborhood has a score of -4.73. The mean socioeconomic status z score of Dutch neighborhoods over the years is zero.
Pleasantness test
Sugar content did not significantly affect whether the children liked the beverages.
Children rated the artificially-sweetened beverages 3.39±0.7 (mean±SD) and the
sugar-sweetened beverages 3.39±0.6 (Table 2.2). The difference was not statistically
significant (P = 0.9). This means that both versions were mostly rated in the categories
‘neutral’ and ‘liked’ (Table 2.3). Also, for each of the seven pairs separately, children
liked the sugar-sweetened and the artificially-sweetened beverages equally. However,
the children preferred some beverages to others, irrespective of whether they were
sugar-sweetened or artificially-sweetened: for example, they neither liked nor disliked
Roosvicee forest fruits, but they liked Spa forest fruits (Table 2.2). Different blends of
sweeteners also did not affect whether the children liked the beverages (Table 2.2).
Triangle test
The children were able to discriminate between sugar-sweetened and artificiallysweetened beverages. In total, the 89 children performed 619 Triangle tests (7
beverages x 89 children; four tests were invalid because the test leader offered
incorrect samples). In 306 (49%) of these tests, the children identified the ‘different’
sample. This is 100 more than the 206 correct responses expected by chance alone (P
= 0.005) (Table 2.4). Also, the children could taste the difference between the two
versions for each of the seven pairs separately, and for both blends of artificial
sweeteners (Table 2.4).
32
SENSORY TASTE STUDY
Table 2.2 Ratings of pleasantness of artificially-sweetened and sugar-sweetened beverages
a
on a 5-point scale as judged by 89 children. We examined artificially-sweetened
compared with sugar-sweetened beverages for all the beverages combined, all the
beverages individually, and all beverages combined with the same blend of artificialsweeteners.
b
Beverage
Artificially-sweetened
Sugar-sweetened
P-value
beverages
beverages
All beverages combined
Mean ±SD pleasantness
Mean ±SD pleasantness
3.4±0.7
3.4±0.6
0.90
Non-commercially
3.7±1.1
3.7±1.1
0.73
custom-made lemon
Non-commercially
3.2±1.1
3.2±1.1
0.84
custom-made mango
Non-commercially
3.3±1.2
3.1±1.2
0.48
custom-made peach
Roosvicee Forest fruits
3.2±1.3
3.1±1.3
0.41
Roosvicee Peach
3.2±1.2
3.2±1.2
0.64
Spa Apple/Cherry
3.5±1.2
3.6±1.0
0.33
Spa Forest fruits
3.7±1.2
3.8±1.0
0.35
All beverages combined
3.3±0.8
3.3±0.8
0.41
with sucralose/acesulfame
c
potassium
All beverages combined
3.5±0.8
3.5±0.7
0.42
with cyclamate/acesulfame
d
potassium/saccharin
a
Pleasantness was rated by a 5-point scale questionnaire with 5 faces representing a
range of liking. We asked the children to point out the face that indicated how much they
liked the beverage.
b
Differences between the types of beverages were analyzed with Wilcoxon signed rank
tests; P ≤ 0.05 was considered to indicate significance.
c
Beverages with sucralose/acesulfame potassium were: Non-commercially custom-made
lemon, Non-commercially custom-made mango, Non-commercially custom-made peach
and Roosvicee Forest fruits
d
Beverages with cyclamate/acesulfame potassium/saccharin were: Roosvicee Peach, Spa
Apple/Cherry, and Spa Forest fruits
33
34
Brand nameb
Flavor
Non-commercially
Lemon
4
10
21
28
26
custom-made
Sugar-sweetened
Non-commercially
Lemon
3
10
24
22
30
89
custom-made
Artificially-sweetened
Non-commercially
Mango
9
9
40
19
12
89
custom-made
Sugar-sweetened
Non-commercially
Mango
5
17
35
19
13
89
custom-made
Artificially-sweetened
Non-commercially
Peach
8
18
23
23
17
89
custom-made
Sugar-sweetened
Non-commercially
Peach
10
13
33
22
11
89
custom-made
Artificially-sweetened
Roosvicee
Forest fruits
12
12
26
24
15
89
Sugar-sweetened
Roosvicee
Forest fruits
14
12
28
23
12
89
Artificially-sweetened
Roosvicee
Peach
7
21
21
24
16
89
Sugar-sweetened
Roosvicee
Peach
8
21
23
24
13
89
Artificially-sweetened
Spa
Apple/Cherry
6
10
26
28
19
89
Sugar-sweetened
Spa
Apple/Cherry
3
10
23
35
18
89
Artificially-sweetened
Spa
Forest fruits
8
8
17
30
26
89
Sugar-sweetened
Spa
Forest fruits
2
5
28
28
26
89
a
Pleasantness was rated by a 5-point scale questionnaire with 5 faces representing a range of liking. We asked the children to point out the face that
indicated how much they liked the beverage.
b
We selected beverages of 3 manufacturers: the Non-commercially custom-made beverage were produced by Unilever (Colworth, U.K.), the Roosvicee
beverages by H.J. Heinz Food Company (Pittsburg, U.S.A.), and Spa beverages by Spadel Group (Brussels, Belgium)
Artificially-sweetened
Type of beverage
Total
number
of
responses
89
Table 2.3 Ratings of pleasantness of the seven artificially-sweetened and seven sugar-sweetened beverages on a 5-point scale as judged by 89 children.a
We measured pleasantness immediately after children tasted the beverages. We present the number of responses for each of the five categories.
CHAPTER 2
SENSORY TASTE STUDY
Table 2.4 Ability of children to discriminate artificially-sweetened and sugar-sweetened
a
beverages.
Beverages
Number of correct
Number (%) of
P-value
responses expected
correct responses
by chance (π=1/3)
observed
All beverages combined
206
306 (49)
0.005
Non-commercially
30
43 (48)
0.000
custom-made lemon
Non-commercially
30
46 (52)
0.002
custom-made mango
Non-commercially
30
54 (61)
0.000
custom-made peach
Roosvicee Forest fruits
30
47 (53)
0.005
Roosvicee Peach
30
41 (46)
0.008
Spa Apple/Cherry
29
32 (37)
0.000
Spa Forest fruits
29
43 (48)
0.000
All beverages combined with
119
190 (53)
0.000
sucralose/acesulfame
b
potassium
All beverages combined with
88
116 (44)
0.000
cyclamate/acesulfame
c
potassium/saccharin
a
We performed one-sided binominal tests because children could or could not identify
the ‘different’ sample. We offered children 3 samples in each triangle test with always a
chance of 33% to pick the correct sample by chance. 30 correct responses or more out of
89 would then lead to a significant P-value at α=5%. For each beverage data were
available for 89 children, except for Spa Apple/Cherry (N=86) and Spa Forest fruits (N=88)
because the test leader offered false samples
b
Beverages with sucralose/acesulfame potassium were: Non-commercially custom-made
lemon, Non-commercially custom-made mango, Non-commercially custom-made peach
and Roosvicee Forest fruits: 4 beverages x 89 children = 356 responses in total, 190 were
identified correct, the rest false.
c
Beverages with cyclamate/acesulfame potassium/saccharin were: Roosvicee Peach, Spa
Apple/Cherry, and Spa Forest fruits: 3 beverages x 89 children – 4 tests that were
invalidated = 263 responses in total, 116 were identified correct, the rest false.
35
CHAPTER 2
Discussion
We found that children liked sugar-sweetened and artificially-sweetened beverages
equally well. Analyses for both the individual pairs and the different blends of artificial
sweeteners did not change this outcome. This result is in line with a previous threeweek study in adults with beverages,7 and with studies of other foods such as
pudding25 or cream cheese26 which also found that sugar-sweetened and artificiallysweetened products were liked equally. However, our study contradicts the results of
a four-week study of beverages which found that beverages sweetened with a blend
of aspartame, acesulfame potassium, and saccharin were rated lower in pleasantness
than beverages containing sugar. Thus the pleasantness of artificially-sweetened
products may be highly dependent on the (mix of) sweeteners used and on other
aspects of product formulation. We found that children liked both versions equally
well, irrespective of the different amounts of sugar added to the artificiallysweetened beverages. Apparently, there is no need to add sugar to artificiallysweetened beverages.
Interestingly, although children liked sugar-sweetened and artificially-sweetened
beverages equally, they were able to discriminate between them. Analyses for both
the individual pairs and the different blends of artificial sweeteners did not change
this outcome. This is remarkable because the artificially-sweetened beverages
contained blends of artificially sweeteners that are known to increase the similarity
with sucrose.15,19,20 However, the artificially-sweetened beverages did not contain
aspartame. Compared with other artificial sweeteners, aspartame is most equivalent
to sucrose.14-18 This may explain the lack of resemblance between the two versions.
This study has several strengths. We managed to single blind the study; during
the tests, the participants did not know whether a beverage was sugar-sweetened or
artificially-sweetened. We also generated unique sequences of the beverages for each
child, which removed the time-order bias. This refers to the situation in which
different orders produce different judgments for the same product.27 One way to deal
with this problem is to provide products in all possible orders, i.e. randomized orders,
so that the sequential effects average out in the group data.28 We also included a
relatively large sample size, when compared with other studies which had 16 to 31
participants.25-27 However, since this study was a pilot study, we did not perform a
power-calculation prior to the study. Finally, we used a validated instrument to
measure liking.21
36
SENSORY TASTE STUDY
Our study also has limitations. We combined beverages with different flavors
and blends of sweeteners to examine the difference preference towards artificiallysweetened and sugar-sweetened beverages. This may be a rather crude approach.
However our data showed that the children liked the beverages with different flavors
and blends of sweeteners almost equally. This finding is in line with previous studies
that showed that adjustments for flavor did not yield different outcomes.16,17 Our
study population included children who habitually consumed sugar-sweetened
beverages during school breaks. A study in adults showed that those who generally
drink sugar-sweetened beverages are better able to distinguish artificially-sweetened
from sugar-sweetened beverages than those who usually consume artificiallysweetened beverages.29 Our study suggests that this ability may also apply to
children. We included children aged between 5 and 12, but most were in the upper
part of this range: there were 56 aged 9 to 12 years, and 30 aged 5 to 8 years. This
may have influenced our findings, because the older children were better at
discriminating between sugar-sweetened and artificially-sweetened beverages than
the younger children (data not shown). Younger children may have experienced more
difficulties with the test despite some practices. Also, we do not know the exact
amounts of artificial sweeteners in the commercial artificially-sweetened beverages.
However, the artificially-sweetened and sugar-sweetened beverages were equally
liked within all the pairs. We therefore do not think that this information is relevant
for our conclusion. Finally, we did not collect baseline characteristics other than age
and gender. We cannot think of any baseline characteristic that may have
strengthened the conclusions, except habitual beverage consumption at home. It may
have been valuable to identify whether children are habitual consumers of sugary or
sugar-free beverages with a parental questionnaire. We now concluded that children
are primarily consumers of sugar-sweetened beverages based on one observation
prior to the study. However, this was only one observation at school and we did not
include any data about the situation at home.
Our participants were healthy Dutch children. Future studies should be carried
out to investigate whether our findings hold for other ethnic groups, obese children,
habitual consumers of artificially-sweetened beverages, and different age groups.
Future studies may also include the paired comparison test to examine whether
participants also like artificially-sweetened and sugar-sweetened beverages equally
when the test leader offers the two beverages at the same time instead of
consecutively. We used a sample of beverages that are similar to beverages currently
available in supermarkets. We therefore speculate that our findings can be
37
CHAPTER 2
generalized to other beverages. However, future studies may examine whether our
findings hold for other types of beverages.
In conclusion, children liked artificially-sweetened beverages and sugarsweetened beverages equally, but were able to discriminate between the two. Recent
trials have shown that replacement of sugar-sweetened beverages by sugar-free
beverages reduces weight gain in children.4,5 Artificially-sweetened beverages may
provide a viable alternative to sugar-sweetened beverages. However, we consider tap
water the preferred option because of the diminished likelihood of causing dental
erosion.30
38
SENSORY TASTE STUDY
References
1. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of obesity and trends in
body mass index among US children and adolescents, 1999-2010. JAMA
2012;307:483-90.
2. Olds T, Maher C, Zumin S, et al. Evidence that the prevalence of childhood
overweight is plateauing: data from nine countries. Int J Pediatr Obes 2011;6:342-60.
3. Wang YC, Bleich SN, Gortmaker SL. Increasing caloric contribution from sugarsweetened beverages and 100% fruit juices among US children and adolescents, 19882004. Pediatrics 2008;121:e1604-14.
4. de Ruyter JC, Olthof MR, Seidell JC, Katan MB. A trial of sugar-free or sugarsweetened beverages and body weight in children. N Engl J Med 2012;367:1397-406.
5. Ebbeling CB, Feldman HA, Chomitz VR, et al. A randomized trial of sugarsweetened beverages and adolescent body weight. N Engl J Med 2012;367:1407-16.
6. Sylvetsky AC, Welsh JA, Brown RJ, Vos MB. Low-calorie sweetener consumption
is increasing in the United States. Am J Clin Nutr 2012;96:640-6.
7. Beridot-Therond ME, Arts I, Fantino M, De La Gueronniere V. Short-term effects
of the flavour of drinks on ingestive behaviours in man. Appetite 1998;31:67-81.
8. Van Wymelbeke V, Beridot-Therond ME, de La Gueronniere V, Fantino M.
Influence of repeated consumption of beverages containing sucrose or intense
sweeteners on food intake. Eur J Clin Nutr 2004;58:154-61.
9. Muir DD, Hunter EA, Williams SAR, Brennan RM. Sensory profiles of commercial
fruit juice drinks: influence of sweetener type. Journal of the Science of Food and
Agriculture 1998;77:559-65.
10. Kamerud JK, Delwiche JF. Individual differences in perceived bitterness predict
liking of sweeteners. Chem Senses 2007;32:803-10.
11. Kuhn C, Bufe B, Winnig M, et al. Bitter taste receptors for saccharin and
acesulfame K. J Neurosci 2004;24:10260-5.
12. Ott DB, Edwards CL, Palmer SJ. Perceived Taste Intensity and Duration of
Nutritive and Non-nutritive Sweeteners in Water using Time-intensity (T-I)
Evaluations. J Food Sci 1991;56:535-542.
13. Ayya N, Lawless HT. Quantitative and qualitative evaluation of high-intensity
sweeteners and sweetener mixtures. Chemical Senses 1992;17:245-59.
14. Schiffman SS, Reilly DA, Clark TB, 3rd. Qualitative differences among sweeteners.
Physiol Behav 1979;23:1-9.
39
CHAPTER 2
15. Schiffman SS, Crofton VA, Beeker TG. Sensory evaluation of soft drinks with
various sweeteners. Physiol Behav 1985;34:369-77.
16. Larson-Powers N, Pangborn RM. Paired comparison an time-intensity
measurements of sensory properties of beverges and gelatins containing sucrose or
synthetic sweeteners. J Food Sci 1978;43:41-6.
17. Larson-Powers N, Pangborn RM. Descriptive Analysis of the Sensory Properties
of Beverages and Gelatins Containing Sucrose or Synthetic Sweeteners. J Food Sci
1978;43:47-51.
18. Cloninger MR, Baldwin RE. Aspartylphenylalanine methyl ester: a low-calorie
sweetener. Science 1970;170:81-2.
19. Hanger LY, Lotz A, Lepeniotis S. Descriptive Profiles of Selected High Intensity
Sweeteners (HIS), HIS Blends, and Sucrose. J Food Sci 1996;61:456-59.
20. Matysiak NL, Noble AC. Comparison of Temporal Perception of Fruitiness in
Model Systems Sweetened with Aspartame, an Aspartame + Acesulfame K Blend, or
Sucrose. J Food Sco 1991;56:823-26
21. Léon F, Couronne T, Marcuz MC, Köster EP. Measuring food liking in children: a
comparison of non verbal methods. Food Quality and Preference 1999;10:93-100.
22. Vlieg-Boerstra BJ, Bijleveld CM, van der Heide S, et al. Development and
validation of challenge materials for double-blind, placebo-controlled food challenges
in children. J Allergy Clin Immunol 2004;113:341-6.
23. Randomization.com. http://www.randomization.com. Accessed at October 10 2008.
24. Sociaal en Cultureel Planburo (English translation: The Netherlands Institue for
Social Research. Socio-economic status score. http://www.scp.nl. Accessed at
December 11, 2012.
25. Rolls BJ, Laster LJ, Summerfelt A. Hunger and food intake following consumption
of low-calorie foods. Appetite 1989;13:115-27.
26. Anton SD, Martin CK, Han H, et al. Effects of stevia, aspartame, and sucrose on
food intake, satiety, and postprandial glucose and insulin levels. Appetite 2010;55:37-43.
27. Amerine MA, Pangborn RM, Roessler EB. Principles of sensory evaluation of
Food. New York, USA: Academic; 1965.
28. Lawless HT, Heymann H. Sensory Evaluation of Food: Principles and Practices.
New York, USA: Chapman &Hall; 1998.
29. Ackerman S, Sullivan E, Walls M. Design and analysis of ROC studies:
Discrimination between diet and regular soft drinks. Academic Radiology
1998;5:S259-S63.
30. Gambon DL, Brand HS, Veerman EC. Dental erosion in the 21st century: what is
happening to nutritional habits and lifestyle in our society? Br Dent J 2012;213:55-7.
40
CHAPTER 3
Effect of sugar-sweetened beverages on body weight in
children: design and baseline characteristics of the
Double-blind, Randomized INtervention study in Kids
Janne C de Ruyter
Margreet R Olthof
Lothar D J Kuijper
Martijn B Katan
Contemp Clin Trials 2012;33: 247–257
CHAPTER 3
Abstract
Background
Intake of sugar-sweetened beverages is associated with overweight in observational
studies. A possible explanation is that liquid sugars do not satiate and that their intake
is not compensated by reduced caloric intake from other foods. However, evidence
from intervention studies for this hypothesis is inconclusive because previous studies
were not blinded. Hence results may have been influenced by expectations and
behavioral cues rather than by physiological mechanisms.
Methods
We designed the Double-blind, Randomized INtervention study in Kids (DRINK) to
examine the effect on body weight of covertly replacing sugar-sweetened by sugarfree beverages. Children were only eligible if they habitually drank sugar-sweetened
beverages. We recruited 642 healthy children (mean age 8.2, range 4.8–11.9). We
designed, tested and produced custom-made beverages containing 10% sugar and
sugar-free beverages with the same sweet taste and look. Children receive one 250
mL can of study beverage daily for 18 months. We perform body measurements at 0,
6, 12 and 18 months. The primary outcome is the z-score of BMI for age. The
maximum predicted difference in this score between groups is 0.72, which
corresponds with a difference in body weight of 2.3 kg.
Discussion
The double-blind design eliminates behavioral factors that affect body weight. If
children gain less body fat when drinking sugar-free than when drinking sugarsweetened beverages that would show that liquid sugar indeed bypasses biological
satiation mechanisms. It would also suggest that a reduction in liquid sugars could
decrease body fat more effectively than reduction of other calorie sources.
44
DESIGN OF THE DRINK TRIAL
Introduction
Obesity in children has become a major health problem worldwide. In the past three
decades the prevalence of overweight in children has increased dramatically.1
Recently the prevalence of high body mass index in children appeared to plateau2 but
the number of overweight children remains high. Obesity in children is a risk factor for
adult obesity, type 2 diabetes, cancer, cardiovascular diseases and death before 55
years of age.3-8 Obesity in children also has negative health consequences during
childhood itself such as insulin resistance, hypertension, dyslipidemia and type 2
diabetes.9 The increase in prevalence of obesity coincided with a large increase in
consumption of sugar-sweetened beverage in children.10 Energy intake among
children increased substantially between 1977 and 2001 in the US, and energy from
sugar-sweetened beverages accounted for more than 50% of this increase.11 Sugarsweetened beverages, also known as liquid sugars, are thought to be more fattening
than sugars in solid form because they do not satiate.12 As a result children would not
compensate for the intake of liquid calories by eating less of other foods and drinks.
Therefore sugar-sweetened beverages might increase total energy intake and cause
overweight.
Indeed, in several observational studies intake of sugar-sweetened beverages
was associated with weight gain both in adults13-15 and in children.16-18 However, some
studies failed to find such an association.19-21 Also, an association between sugarsweetened beverages and weight gain may be due to confounding because people
who drink more soft drinks often also eat more fast food and exercise less. Although
most studies adjust statistically for known confounders, residual and unmeasured
confounding are of increasing concern in observational nutrition studies.22,23 Hence
trials are needed for conclusive evidence about causality.
Results of trials of the effect of liquid sugars on overweight are inconclusive,
with some finding an increase in weight24-27 while other trials found equivocal
results.28-30 The lack of conclusive evidence may be due to a lack of proper placebo
treatments, small samples sizes, short duration and lack of individual randomization.
Also, none of these trials was properly blinded. Blinding is crucial because without it,
results may be biased by effects on food intake of behavioral cues and expectations
rather than by physiological mechanisms.31 Hence there is a need for conclusive
evidence whether liquid sugars fail to evoke compensatory changes in food intake and
are therefore more fattening than other sources of calories.32,33 We designed DRINK –
the Double-blind, Randomized INtervention study in Kids – to examine the effect of
45
CHAPTER 3
covertly replacing sugar-sweetened beverages by sugar-free beverages on
bodyweight. This allows us to study biological compensatory mechanisms
independent of behavioral cues and voluntary changes in intake. In this article we
describe the rationale, design, methods and baseline characteristics of this study.
Methods
Outline
DRINK is an individually randomized, double-blind, controlled, parallel intervention
study in free-living school children. For 18 months, 642 school children randomly
receive daily either one can of sugar-free beverage (treatment) or one can of sugarsweetened beverage (control). We would have preferred to use water as the
treatment beverage, but that would make blinding impossible. Children consume the
beverage at school on weekdays during their morning break, and at home during
weekends and holidays. The study is double-blind; neither the researchers nor the
children, parents or teachers know who drinks which beverage. The study is
conducted in the Netherlands at eight schools in an urbanized area near Amsterdam.
We distribute the beverages to the schools (bi)weekly by courier. The study was
preceded by a pilot study to test feasibility and logistics and by a sensory study (see
Formulations) to help us in developing the beverages.
DRINK pilot study
The pilot was a two-month study conducted in April–May 2009 at one elementary
school in Purmerend, 20 km from Amsterdam. The study was conducted to the
guidelines laid down in the Declaration of Helsinki and the study protocol was
approved by the Medical Ethical Committee of VU University Medical Centre
Amsterdam. Written informed consent was obtained from a parent or guardian. In
the rest of this paper ‘parent’ will refer to both parents and guardians. The inclusion
and exclusion criteria for the pilot were the same as for the main study (see
Recruitment). Forty-four children aged 5–12 were randomized to receive daily one
330 mL can with either sugar-sweetened or sugar-free beverage (bought from
Unilever, Colworth, UK). Both were available with lemon and peach flavor. Our courier
transported the beverages to the school.
We asked various stakeholders for advice on how to set up the main DRINK
study. Frequent conversations with the children taught us that offering more than
two flavors would help to increase compliance, and that newsletters, contests and
46
DESIGN OF THE DRINK TRIAL
birthday cards would be appreciated. Therefore we implemented these and other
forms of entertainment in the main study. The pilot lacked entertainment except the
possibility to collect fancy stickers that we attached to the cans in the last two weeks.
In an evaluation form teachers reported that it was not a problem for them to
accommodate consumption of our beverages at school. The parents reported that the
children liked the pilot and that they were keen on bringing home the weekend and
holiday cans. Parents also brought up that it would be impossible to transport cans
during the summer holidays of 6 weeks. Therefore we interrupted treatment in the
main DRINK study for 6 weeks in the summer of 2010, and prolonged it by 6 weeks at
the end. Parents also did not like to return empty cans used at home back to school
during the pilot. Hence we collect cans used at school but not cans used at home to
assess compliance during the main study (see Ancillary measurements). Our advisory
board suggested that schools might expect something in return for their participation.
Therefore we offered them health related education materials. Two schools made use
of this option.
Design and setting of main DRINK study
The study protocol was approved by the Medical Ethical Committee mentioned
above. DRINK is an individually randomized, double-blind, controlled, parallel
intervention study in free-living school children aged 5–11 years (Fig. 3.1). All children
receive 250 mL of the study beverage per day which provides either 0 g or 25 g of
sugar. Children in the control group do not change their liquid caloric intake because
they replace their customary sugar-sweetened beverage brought from home by a
sugar-sweetened study beverage. Children in the treatment group replace their
customary sugar-sweetened beverage by a sugar-free study beverage. Therefore they
consume 100 liquid calories per day less than the control group. We conduct the
study at eight elementary schools, in the Zaanstreek, Purmerend and Haarlem, in an
urbanized area, 16–33 km from Amsterdam. It is customary for children in Dutch
elementary schools to consume a beverage brought from home in class during a
morning break around 10 am under supervision of the teacher. We replace the
beverage brought from home by a study beverage and ask the teacher to check if
children consume their beverage during the morning break and to remind them to
take home weekend and holiday cans. Parents supervise consumption of the study
beverage on weekend days and during holidays.
47
CHAPTER 3
4913 children approached
(22 elementary schools)
628 children refused (13%)
2850 children no answer (58%)
1435 informed consents (29%)
14 schools with 736 consenting children excluded
• 8 schools with 392 consenting children due to insufficient number of
informed consents per school
• 6 schools with 344 consenting children due to less accessible school
location
699 informed consents
(8 elementary schools)
57 children not eligible
• 5 too young
• 2 on treatment for obesitas
• 12 insufficient habitual consumption of sugar-sweetened beverages
• 2 chronic diseases
• 32 late refusals
• 4 others
642 children randomized
Treatment group
(321 children)
Control group
(321 children)
1 can
per day
of 250 mL
sugar-free beverage
1 can
per day
of 250 mL
sugar-sweetened
beverage
Figure 3.1 Flow of recruitment and design of the Double-blind Randomized INtervention study in Kids (DRINK) that studies the effect of sugar -sweetened beverages
on body weight.
Beverages
Formulations
The development and manufacture of the beverages was the major challenge for our
study. We needed pairs of beverages, one sugar-free and one sugar-sweetened,
which tasted and looked the same. The beverages needed to be safe and stable,
attractive and pleasant-tasting for children. They also had to be acceptable for
parents and schools. We opted for a non-carbonated beverage because of the poor
health image of fizzy drinks. We offered several flavors – raspberry, lemon, peach and
mango – to increase compliance.
Before the study started we tested the sensory qualities of these beverages in 89
children aged 5–12 in Purmerend at the same school where we did the pilot study.
The children liked the sugar-free beverages and sugar-sweetened beverages equally.
The sugar-sweetened beverages contain 25 g of sugar and 100 kcal per 250-mL
can (Table 3.1). The sugar-free beverages do not provide calories. They contain the
artificial sweeteners sucralose (0.157 g/L) and Acesulfame-K (0.057 g/L). These
sweeteners are approved by the US Food and Drug Administration, the Joint
Commission of Experts on Food Additives of the World Health Organization, the Food
48
DESIGN OF THE DRINK TRIAL
and Agriculture Organization, and the European Food Safety Authority. The
acceptable daily intake (ADI) for sucralose is 15 mg/kg bodyweight/ day.34 This means
that a child of 20 kg can drink seven cans of sugar-free study beverage per day. The
ADI for acesulfame-K is 9 mg/kg bodyweight/day 35 which corresponds with 13 cans.
Since we offer one can per day, the amount of artificial sweeteners is safe.
We use vegetable and fruit concentrates as colorings in the mango, peach and
raspberry flavored beverages. The lemon drinks do not contain colorings.
Table 3.1 Composition of study beverages in the Double-blind, Randomized Intervention
a,b
study in Kids (main DRINK study).
Sweetener
Calories
Sugar-sweetened beverage
per 250-mL can
Sugar 25 g
100
Sugar-free beverage
per 250-mL can
Sucralose 0.04 g
0
Acesulfame-K 0.01 g
a
Ingredients besides the sweetener: Fruit flavor (lemon, mango, raspberry or peach), fruit
juices (lemon, mango, raspberry or peach), colorings ( GNT Exberry® vegetable and fruit
concentrates; no color in lemon), water, trisodium citrate, citric acid, ascorbic acid, malic
acid (malic acid only in sugar-sweetened beverage).
b
We bought the colorings from GNT (Mierlo, The Netherlands) and flavorings from
Unilever (Colworth, UK). All other ingredients were from Refresco Benelux (Maarheeze,
Netherlands).
Production
We hired JJM (Brussels, Belgium) to design the cans, Crown (Wantage, United
Kingdom) to make the cans, and Refresco Benelux (Maarheeze, Netherlands) to
produce the beverages and fill the cans. Fig. 3.2 presents the can design. We used 250
mL cans because that fitted better with the factory's logistics than 330 mL cans and
because the shelf life of cans is better than that of cardboard containers. Each of the
four flavors was produced as a sugar-sweetened and a sugar-free variant. The first
batch consisted of 30000 cans per flavor/sweetener combination for a total of 240000
cans, plus 20000 cans of sugar-sweetened peach beverage for the run in period. These
were produced on 24 August 2009, prior to the start of the study. The second batch of
200000 cans, i.e. 25000 per flavor/sweetener combination, was produced on 7
September 2010. The beverages will never be available commercially. VU University
49
CHAPTER 3
has registered the brand name (“Blikkie”, colloquial for “little can”) and the can design
to prevent commercial exploitation by third parties.
Figure 3.2 Can design of the sugar-free and sugar-sweetened beverages. Blikkie is
colloquial Dutch for “little can”. Telephone number and email address of the research
team and logo of the study and VU University are depicted on the right side of the
can. Expiration date is imprinted at the bottom of each can.
Sample size calculation
The sample size calculation was based on the expected effect of treatment on the
change of the z-score BMI for age and its SD. BMI stands for Body Mass Index and is
the weight divided by height squared (kg/m2). The z-score is the number of standard
deviations by which a child differs from the mean BMI of children of the same age and
gender. We calculated the number of children required as N=7.9×2×(within person)SD
of the change in z-score with time/expected difference of change in z-score)²
(power=0.8; α=0.05).36
We calculated the expected difference of the change in weight and z-score
between treatments from the difference of sugar intake. Children in the treatment
group reduce their intake of liquid sugar by 25 g per day which equals 12600 g of
sugar or 50400 kcal in 18 months. A naive calculation would suggest that at 7709 kcal
per kg weight change37 this would equate a difference of 6.54 kg. However, any
reduction in weight reduces the basal metabolic rate and the energy costs of
activity.38 Therefore we predict that the mean difference of change in weight between
50
DESIGN OF THE DRINK TRIAL
the treatment and control group is 2.3 kg, if children receiving sugar-free drinks do
not compensate for the absence of sugar by eating and drinking more of other
foods.39
We converted the predicted difference of change in weight into a BMI for age zscore because in children the relation between BMI and body fatness depends on age
and gender, while the BMI for age z-score does not. Hence the BMI for age z-score or
standardized BMI allows comparison of children of different ages and gender.
We estimated the difference of change in BMI for age z-score from the expected
difference in weight change. At the end of the study our subjects will on average be
aged 9.5 years. The mean weight of Dutch children aged 9.5 years is 32.25 kg, their
mean height is 1.408 m and mean BMI 16.24 kg/m2.40 A weight reduction of 2.3 kg will
produce a weight of 29.95 kg and a BMI of 15.11 kg/m2. Thus the maximum expected
effect of the treatment is a reduction of 1.13 kg/m2. In Dutch children of this age, a
fall of 1.13 kg/m2 in BMI will shift a child downwards in the distribution by 0.72
standard deviations.40
The effect of removing sugary drinks on health could still be relevant if only part
of the sugar removed was compensated by increased intake of calories from other
sources. Also, we cannot expect 100% compliance to treatment. We designed the
study so that we would still pick up a difference if children in the sugar-free group
compensated 50% of the calories lost from sugar by increased intake of other calorie
sources, and if all participants only consumed 50% of the drinks. The predicted
difference of change in z-score then becomes 50%×50%×0.72=0.18.
We calculated the within-person SD of the change in z score from the changes in
height and weight of 1017 participants in the ChecKid study.41 ChecKid examines
trends in overweight and obesity among children in Zwolle initially aged 4–12 years.
Using raw data obtained from children when they were aged 6.7 and 9.7 we
calculated that the change in BMI for age z-score over three years, relative to the
Dutch mean averaged 0.03 and its SD equaled 0.66.
Hence our sample size calculation is: N=7.9×2×(0.66/0.18)²=212 subjects per
treatment. We recruited 642 because we anticipated a dropout of one third.
Recruitment
We recruited participants between August and November 2009 (Fig. 3.1). We first
approached the boards of seven school districts that encompassed 131 schools for
permission to contact their schools. We approached another 11 schools directly. Out
of these 142 schools, 22 were willing and able to cooperate and granted us access to
51
CHAPTER 3
parents and children. The 4913 pupils at these 22 schools received a recruitment flyer
including a consent form. We set up information booths at the schools to tell parents
and children about the study. We also let them taste the study beverages and we sent
out a press release which led to broad exposure in the media. An independent
physician was available for parents to answer health related queries. Parents of 1435
children (29%) gave written informed consent for their child to participate. We asked
children to give verbal assent to the parent. Parents that gave their consent were
asked to complete a questionnaire so we could determine if the child was eligible. We
considered it unethical to increase the children's sugar intake. Therefore an important
inclusion criterion was that children had to habitually consume sugar-sweetened
beverages during the morning break at school on at least three out of five school
days. Other criteria were:
Inclusion criteria:
• Habitual consumption of 250 mL or more per day of sugar-sweetened beverages on
at least three out of five school days
• Minimum age of 5 years at the start of DRINK
• Young enough to be still in elementary school at the end of DRINK
• Written informed consent by a parent
Exclusion criteria:
• Medication or medical treatment for obesity
• Diabetes, growth disorders, celiac disease, or serious gastro-enterologic diseases,
e.g. inflammatory bowel disease
• Medical history or surgery known to interfere with the study
• Participation in another intervention study up to 3 months before and during DRINK
if that interfered with our study
• Physical disabilities that interfere with the measurements
• Plans to relocate and change school during DRINK
Because 1435 children volunteered and we only needed 640, we excluded schools
that were difficult to reach or had few prospective participants. This left 699
participants. Application of the inclusion and exclusion criteria left 642 children to be
enrolled and randomized (Fig. 3.1).
The questionnaire also contained demographic and ethnic questions about the
child and its parents. A child is considered Dutch if both parents are born in the
Netherlands, Non-western if one or both parents are born in Suriname, Dutch Antilles,
52
DESIGN OF THE DRINK TRIAL
Turkey or Morocco, or Other if both parents are born in a country other than these. In
case of a single parent household, we used the country of birth of this parent. We
determined educational level based on both parents, whichever was highest. Level of
education was graded as: Elementary school, Lower vocational secondary education
or technical secondary education, Intermediate secondary education, Intermediate
vocational education, High-school graduate and Higher vocational education/college
degree. We used international cut-offs for overweight and obesity in children and for
low and healthy BMI.42,43
Baseline characteristics
Most of our 642 participants have a normal weight (Table 3.2). Their mean BMI for
age z-score is 0.03 where 0 is the Dutch average in 2009. The number of non-western
children in our population is slightly higher than that in the Dutch population as a
whole.44 In 44% of the households the highest education level attained by either
parent is Intermediate vocational education or less, and 55% had a High-school or
college degree. The level of education is slightly higher than the Dutch average.45
Randomization of participants and beverages
Randomization was done by a statistician (L.D.J.K.) who is not involved in the
execution of the study. Eligible children were individually randomized. An Excel visual
basic macro program randomly assigned children to sugar-sweetened or sugar-free
beverages within each school so that mean age, gender and initial BMI were equal
between treatments. Randomization was done with minimal human intervention, as
follows. Data of children who met all inclusion and exclusion criteria were closed out
in the central database (MS Access) and exported per school into an MS Excel
workbook. The Excel macro written by L.D.J.K. assigned all children a randomization ID
number. Children in the same household received the same randomization ID number
and thus the same treatment, because there is a risk that they drink each other's
study beverages at home. Children, parents and teachers have not been informed
that siblings receive the same treatment.
The macro then assigned to each randomization ID number a random number
sampled from a continuous uniform distribution between 0 and 1, e.g. 0.64451. The
macro sorted children by this random number, and the upper half was assigned to
one treatment and the lower half to the other treatment. The program then
calculated the following percentages per treatment group: males; birth year 1998–
1999; birth year 2000–2001; birth year 2002–2004; BMI<15; BMI 15–18 and BMI>18.
53
CHAPTER 3
Table 3.2 Baseline characteristics of the 642 children in the Double-blind Randomized
INtervention study in Kids (DRINK).
Characteristics
n (%) or Mean ±SD
No. of children
642
Girls
301 (46.9%)
Age (years)
8.2 ± 1.9
Ethnicitya
Dutch
502 (78.2%)
Non-western
120 (18.7%)
Other
12 (1.9%)
b
Household highest educational level (Dutch translation)
Elementary (primary) school (Lagere school / basisonderwijs)
6 (0.9%)
Lower vocational secondary education or technical secondary
30 (4.7%)
education (LBO, LTS, LEAO, Lagere tuinbouw)
Intermediate secondary education (MAVO / MULO)
51 (7.9%)
Intermediate vocational education (MBO, MBA, LO-akten, MTS, MEAO)
195 (30.3%)
High-school (HAVO / VWO, MMS, HBS)
92 (14.3%)
Higher vocational education/ college degree (HBO/ Universiteit)
259 (40.5%)
Weight (kg)
30.19 ± 8.86
Height (cm)
132.5 ± 12.6
BMI
16.8 ± 2.6
BMI for age z-score
0.03 ± 1.04
e
Weight status
Low BMI
5 (0.8%)
Healthy BMI
514 (80.0%)
Overweight
101 (15.7%)
Obese
22 (3.4%)
Skinfold thicknessc
Biceps (mm)
6.8 ± 3.2
Triceps (mm)
11.8 ± 5.0
Subscapular (mm)
8.2 ± 4.8
Supra-iliac (mm)
9.2 ± 5.8
Sum of skinfolds (mm)
36.0 ± 17.8
Waist-to-height ratio (%)
44.4 ± 4.0
d
Electrical-impedance fat mass
(kg)
5.74± 3.75
(%)
17.83± 6.87
a
N = 634; Eight households refused;
b
N = 633; Nine households refused;
c
N = 641; One child refused;
d
N = 638; Four children refused;
e
We used international cut-offs for overweight and obesity in children and for low and healthy BMI.42,43
54
DESIGN OF THE DRINK TRIAL
Ideally all these percentages should come out 50/ 50. If any of these percentages was
<45% or >55% the macro automatically assigned new random numbers and reiterated
the process until gender, birth year and initial BMI were evenly divided between the
treatments, i.e. within each school each treatment group contained more than 45%
and less than 55% of each gender, birth year category and initial BMI category. The
macro ran from start to finish without human intervention. It reported the final
randomization assignments, the percent distributions of gender, birth year and BMI
over treatments, and the number of iterations needed to achieve this. The statistician
assigned the upper and lower half of the random number distribution to intervention
or control treatment by a one-time flip of a coin for the first school randomized; this
assignment then held for all schools. When a school had an odd number of
participating households the assignment of the median child or household alternated
from one school to the next. This procedure is equivalent to blocked randomization,46
but its automation is more straightforward and transparent. The statistician sent the
treatment codes to the database manager who entered them into the central
administration database. The database manager is not further involved in the study.
L.D.J.K also produced anonymous codes for the beverages. He randomly
assigned the codes 1001 to 1008 to the eight types of beverages (4 sugar-sweetened,
4 sugar-free) and sent these to the factory. Beverages for the run-in period (all sugar
sweetened) received code 1009. The factory printed the beverages codes on the
cardboard trays on which cans were packed. The bottom of each can was imprinted
with a 17-digit code into which the beverage code was encrypted. This encryption
produced 100 different can codes per beverage type, for a total of 800 can codes.
Personnel at the warehouse where cans are weekly repackaged for the children
are inevitably aware which 4 codes together form one treatment group, but do not
know what the treatment is. When cans are repackaged they are separated from their
cardboard trays, and the codes on the bottom of the cans are almost always different
between cans even if treatments are the same. Only our statistician can track down
the treatment code for an individual can. So for parents, teachers, children and study
researchers the codes on individual cans cannot reveal the treatment.
Treatment assignments and beverage codes are held by L.D.J.K. and by the
outside database manager. The University's attorney holds a print of treatment
assignments and beverage codes in a sealed envelope.
55
CHAPTER 3
Packaging and distribution of study cans
The cans were shipped from the factory to the warehouse of our courier in Egmond
aan den Hoef, 50 km northwest of Amsterdam, and stored at ambient temperature.
Between April and July 2011 the beverages will be stored at 18 °C to prevent
deterioration of taste.
At the warehouse, one researcher coordinates the repackaging of the cans
according to a standardized procedure. We package the cans per week per child into
11×21×15.5 cm cardboard ‘kid boxes’ imprinted with the study logo. A kid box holds 8
cans: 5 for school days, 2 for weekend days and 1 spare can. Each kid box also
contains 8 name labels that children stick onto their cans each Monday morning to
avoid mix-ups in class. The kid boxes are delivered to the schools in 35×44×16 cm
cardboard ‘class boxes’, which hold up to 6 kid boxes. Our courier distributes the
boxes bi weekly – or every week if the school lacks storage space – and brings back
empty and unused cans. We distributed extra spare cans to all children at the
beginning of the study to be used in case children forget to bring home their weekend
and/or holiday cans. After the second production in September 2010 we distributed
new spare cans.
Body measurements
Measurements are done according to a standardized protocol at t=0, 6, 12 and 18
months. Children are measured in their underwear during school hours, generally
between 8.30 am and 3.30 pm. We ask children to visit the bathroom before the
measurements. Children who stopped consuming the study beverages continue to be
measured if the parent and the child permit us to do so.
Two female researchers are always present simultaneously at each
measurement as chaperones. We measure bodyweight to the nearest 0.1 kg on a
Marsden MPMS-250 digital scale (Oxfordshire, United Kingdom). We calibrate the
scale each measurement day with a KERN 366-98 standard weight of 20 kg (Balingen,
Germany). We measure height to the nearest millimeter with a SECA 214 (Hamburg,
Germany). Height will be measured in duplicate at 18 months. We measure waist
circumference twice to the nearest 0.1 cm at the midpoint between the bottom rib
and the top of the hipbone with a SECA 201 flexible steel tape measure (Hamburg,
Germany).47 We measure biceps, triceps, subscapular, and supra-iliac skinfolds in
triplicate with a Harpender Skinfold Caliper HSK-BI (Burgess Hill, United Kingdom).47
Two specially trained researchers perform the waist circumference and skinfold
measurements. Each child is measured by the same researcher throughout the study
56
DESIGN OF THE DRINK TRIAL
to reduce variation. We measure arm-to-leg electrical impedance twice with a
BodyStat 1500MDD (Douglas, United Kingdom). We put 3M 2330 electrodes (St. Paul,
U.S.A.) on the right hand and foot according to the manufacturer's manual. Then we
connect the two cable leads of the BodyStat to the electrodes and perform the
measurements according to standardized procedures.48 We calibrate the BodyStat
each measurement day with the BodyStat calibrator. The calibrator is a small device
onto which we connect to the cable leads to perform a measurement. The calibration
has succeeded if the impedance is between 496 and 503 Ω.
Ancillary measurements
We measure compliance by counting returned cans and by analysis of sucralose in
urine. We count cans from school days only, because parents do not return cans used
at home during weekends and holidays. On school days children place their empty
cans back into their kid boxes, and these are collected by our courier together with
the unused cans. We check returned cans one week each month i.e. 25% of all cans
returned. Cans are scored as empty, half-filled or full.
We will measure sucralose in urine of the children in the sugar-free group. A
study in humans showed that 14.5% of ingested sucralose is excreted in urine.49
Therefore we can use sucralose in urine as a compliance marker. We will also measure
urine samples from children in the sugar-sweetened group as a control. Compliance is
assumed to be similar in both groups since the sugar-sweetened and the sugar-free
beverages were equally appreciated by the children (see Formulations). Therefore we
collect spot urine samples from all children at t=0, 6, 12 and 18 months.
We administer a short dental questionnaire and a hedonic questionnaire at t=12
and t=18 months. The dental questionnaire inquires about the number of new dental
fillings and teeth newly extracted because of caries. The hedonic questionnaire asks:
1. How much do you want to drink the study drink? 2. How satiated do you feel? 3.
What do you eat together with the study drink? 4. How much do you like the study
drink?
Endpoints
Our null hypothesis is that children in the sugar-free group will fully compensate for
the loss of the sugar from their habitual drinks by increasing their intake of calories
from other foods and beverages. Under this hypothesis we will find no difference
between the sugar-free and sugar-sweetened group in the change in BMI for age zscore between t=0 and t=18 months. The alternative hypothesis is that compensation
57
CHAPTER 3
is not 100% and that children receiving sugar-free beverage will gain less body fat
than those in the control group.
The primary endpoint is the difference between the sugar-free and sugarsweetened group in the change in BMI for age z-score between 0 and 18 months. The
secondary outcomes are the waist-to-height ratio (%), the sum of the four skinfolds
(mm) and body fat percentage estimated from electrical impedance (kg and %).
Waist-to-height ratio is an accurate measure of body fat in children that does not
require sex and age specific adjustments.50 Skinfold thickness is also an accurate
indicator of body fat in children.51 We will estimate fat percentage from the electrical
impedance as follows:52 Body fat (kg)=body weight−fat free mass. Fat free mass
(kg)=0.622×height2/R50 (Ohm)+0.234×weight (kg)+1.166×R50=resistance at 50 kHz.
Data analysis and statistics
We consider a per-protocol analysis most suitable because our research question
examines the biological effect of sugar-free beverages on weight instead of the
effectiveness to implement an intervention of sugar-free beverages. We will also do
an intention-to-treat analysis to take into account selective dropout and to prevent
ignoring children who stopped due to effects of the study beverages on body weight.
Our primary analysis will thus be per-protocol analyses of those children who
consumed the study beverages until the end of the study. The primary endpoint is
BMI for age zscore, and the secondary endpoints are sum of four skinfolds (mm),
waist to height ratio (%) and fat mass determined from electrical impedance (kg and
%). The difference in change from 0 to 18 months between the sugar-free and sugarsweetened group will be analyzed with the independent samples t-test. For each
endpoint we will report the mean difference in change between the sugar-free and
sugar group and its 95% confidence interval and p-value. To check the robustness of
this analysis we will also perform a linear regression analysis that corrects for the
outcome variables at baseline. We do not expect that this will lead to markedly
different results because we randomized the participants into two large groups and
stratified for school, gender, age and initial BMI.
Approximately a third of the participants are siblings. We will perform subgroupanalyses in which we will collapse siblings from the same household into one fictional
subject who will be assigned the mean value of these siblings. This will reduce the
number of subjects and eliminate effects of interdependence between siblings.
We will also perform subgroup-analyses for the 80% most compliant children,
where we exclude the 20% of the subjects with the lowest empty can counts.
58
DESIGN OF THE DRINK TRIAL
Secondary analyses will be intention-to-treat analyses. These analyses will
include children who stopped drinking the lemonades prematurely but were available
for the final measurements at 18 months. We expect that final measurements will be
available for 613 out of 642 children randomized. We will not use the “last
observation carried forward” method or imputation techniques because there are
many objections against these methods, especially in growing children. We will
perform the same analyses on this dataset as described above for the primary
analyses.
The second and third timepoints, t=6 and t=12 months, will be used for graphs
but not for analyses. The level of significance will be P<0.05 (2-tailed). Treatment
codes will be broken after blinded data-analysis. For the blinded data-analysis we will
use the numbers 0 and 1 for the treatment groups. After the analyses we will ask our
statistician (L.D.J.K.) to deblind the dataset.We will use the Statistical Package for the
Social Sciences (SPSS) version 17.0 to perform the analyses.
Timelines
Fig. 3.3 presents the planning of the study. After the recruitment and baseline
measurements we randomized the children into the two treatment groups. There
were no dropouts between the baseline measurements and randomization. Children
started the study with a run-in period of one week in which they were all given the
sugar-sweetened peach lemonade. This run-in period served to iron out remaining
logistic problems and did not contribute to the length of the intervention period. The
first two schools started the intervention on November 14, 2009, another three
schools on November 30, 2009 and the final three schools on December 7, 2009. The
last three schools started without a run-in period because of time restrictions. We
interrupted treatment during the summer holidays of 6 weeks from July 10, 2010–
August 22, 2010. Final measurements are done after 18 months intervention and the
study therefore ended in July, 2011.
59
CHAPTER 3
Start treatment
Nov-Dec 2009
End treatment
July 2011
Start run-in
Nov 2009
Randomization
Okt-Nov 2009
Measurements
month 6
April-May 2010
Measurements
month 12
Nov-Dec 2010
Measurements
month 18
June-July 2011
Baseline
measurements
Okt-Nov 2009
Recruitment
Aug-Nov 2009
2009
2010
2011
Figure 3.3 Study planning of the Double-blind Randomized INtervention study in Kids
(DRINK). Body measurements and sensory evaluations are performed at t = 0, 6, 12
and 18 months, dental health questionnaires at t = 12 and 18 months.
Discussion
Low-caloric drinks seem a healthy alternative to sugar-sweetened beverages to
prevent obesity in children because sugar-sweetened drinks may fail to satiate and
therefore add calories on top of the rest of the diet. However, conclusive evidence
that sugar-sweetened beverages are more prone than other foods to cause
overweight is lacking. Therefore we designed the DRINK study. Our study examines if
children who are switched to sugar-free beverages increase their caloric intake from
other sources to compensate the caloric deficit in these beverages, or not.
Strengths and limitations
Our study has several strengths. First, the study is double-blind; nobody knows if a
child is drinking sugar-sweetened or sugar-free beverages. Therefore children in the
sugar-free group receive no explicit or subconscious cues that they are expected to
lose weight, and children in the sugary group are not encouraged to eat less or
exercise more so as to avoid weight gain. Hence this study will be a strict test of
effects of liquid sugars on body weight through unconscious physiological
60
DESIGN OF THE DRINK TRIAL
mechanisms that regulate food intake. Previous trials in children28-30 and in adults27
on the relation between sugar-sweetened beverages and body weight investigated
the effects of educational programs or environmental interventions. These trials were
not blinded and could not separate changes in body weight due to behavioral changes
from changes due to physiological factors. Other trials in adults24-26 on the relation
between sugar-sweetened beverages and body weight were also not properly
blinded.
A second strength is that treatment lasts 18 months and that we have 642
participants. Other trials24-26 were relatively short, varying from 3 to 10 weeks, and
sample sizes ranged from 15 to 42 subjects.
A third strength is the use of a proper placebo treatment. A study on the effect
of sugar-sweetened beverages on weight gain requires sugar-sweetened and sugarfree drinks that taste and look the same. We managed to develop and produce these
beverages. Earlier trials used other treatments. Tordoff25 used high-fructose corn
syrup versus aspartame sweetened beverages, Raben26 used sucrose-sweetened
drinks and foods versus artificially-sweetened drinks and foods, DiMeglio24 used jelly
beans versus caffeine-free soda.
The fourth strength is the individual randomization, which will average out other
behaviors that affect weight. For example, both groups should have similar numbers
of children that watch a lot of television or change their TV watching habits, that eat a
lot of candy, or that become more or less active. Also, parental education and
ethnicity are similar for both experimental conditions. Furthermore, we stratified
participants by school, gender, age and BMI category to ensure that these potential
determinants of weight gain were evenly distributed between the treatment and
control group. Therefore we assume that any difference in change in endpoints can be
attributed to the study beverages and not to other life style factors.
Our study also has limitations. First, the hypothesis that we test deals with
satiation and food intake, and ideally we should measure what the children eat and
drink in response to the treatments. However, it is impossible to measure food intake
in free living subjects to the degree of precision required for our study. As body
fatness is the outcome of concern, we decided to concentrate on this endpoint.
Follow-up experiments with smaller groups that examine the underlying mechanisms
of the association between liquid carbohydrates and weight status might consider
measurements of caloric intake.
Second, the ideal endpoint would have been fat mass estimated from dualenergy X-ray scans. This was not feasible. However, if the study beverages cause
61
CHAPTER 3
differences in BMI for age z-score it is plausible that these are due to differences in fat
mass, and we have other body measurements to back this up.
Third, our primary compliance marker is counts of returned empty cans that
children consumed during school days. We did not collect empty cans from the
weekends and holidays. In the sugar-free group analysis of sucralose in urine will
provide a measure of compliance that is not available for the children receiving sugarsweetened lemonade. However, our objective measurements of compliance still
exceed those in most other trials of diet and body weight.
Fourth, we excluded schools that were difficult to reach or had few prospective
participants. We do not think that this has an effect on the generalizability of the
results. We recruited schools from an urbanized area near Amsterdam. We included
the children from six towns and excluded children from six towns that partly
overlapped. Based on zip codes of schools included and excluded, socioeconomic
status is quite similar and close to the average Dutch socioeconomic status. In
addition we do not study behavior but a physiological mechanism that in principle
should not be much sensitive to socioeconomic status or place of residence.
The study population consists of healthy children and focuses on biological
satiation mechanisms that may be widely shared. We therefore assume that findings
of this study can be extrapolated beyond the healthy Dutch children of our study. On
the other hand the effect of sugar-sweetened drinks on body weight might depend on
body size, genetics, ethnic factors or health, and thus our findings should be
extrapolated with caution and confirmed in other populations.
Interpretation of outcomes
There are two possible outcomes of our study. If we find no difference in weight gain
between the sugar-free and sugar-sweetened group, then children in the sugar-free
group must have increased their caloric intake from other sources to compensate for
the calories lacking from the study beverage. An exclusive focus on sugar-sweetened
beverages might then be less effective to decrease overweight in children. Other
interventions than reducing intake of sugar-sweetened beverages to decrease
overweight in children should receive equal emphasis. Still, a lower intake of sugarsweetened beverages would help to prevent weight gain because they are a major
source of calories.
If both treatments have the same effect on weight it could also be argued that
the artificial sweeteners in the sugar-free beverages stimulated caloric intake.53 Some
animal studies supported this hypothesis 54-56 but experiments in humans 57-59 did not.
62
DESIGN OF THE DRINK TRIAL
The hypothesis that artificial sweeteners increase energy intake would therefore not
be the most likely explanation if we find no difference between the groups.
If we find that weight gain slows after removal of liquid sugar then children in
the sugar-free group evidently did not increase their intake of other foods and drinks
sufficiently to compensate their reduced intake of liquid calories. It would also
suggest that the converse holds true, i.e. that adding liquid sugar to the diet causes
weight gain because it is not spontaneously compensated by reduced intake of other
foods. That would corroborate the liquid sugar theory and suggest a unique role for
liquid sugars in causing weight gain in children. This outcome would further support
current recommendation to discourage sugar-sweetened drinks and promote noncaloric drinks such as water to prevent weight gain in children. However, one piece of
evidence would still be missing: does covert removal of calories from solid foods also
lead to weight reduction? Current theory suggests that it would not, and that such
calories would be compensated for from other foods. However, only a double blind
trial similar to ours can provide decisive evidence on this reduction.
Conflict of interests
The authors declare that they have no conflict of interest.
Author contributions
J.C.R. coordinated the pilot study, sensory study and the main trial. She also
supervised the data collection, co-obtained medical ethical approval, supervised data
management, was responsible for the recruitment of subjects, managed logistics of
the trial and drafted the manuscript. M.R.O. co-obtained funding, obtained medical
ethical approval, contributed to the design, co-supervised the study and co-supervised
the development and manufacture of the beverages. L.D.J.K. randomized the
participants and performed the coding of the beverages. M.B.K supervised the study,
conceived and designed the study, obtained funding and supervised the development
and manufacture of the beverages. M.R.O, L.D.J.K and M.B.K. revised the manuscript
critically.
Acknowledgments
We thank the staff, teachers, parents and children of the schools for their willingness
to participate and the pleasant cooperation during the study; Emilie de Zoete and
Hetty Geerars for excellent assistance in the execution of this study; Mr. Joop Bremer,
retired head of research of a Dutch soft drink manufacturer, for his help and advice
63
CHAPTER 3
during development and manufacture of the beverages; Refresco Benelux for
meticulous care in producing the beverages; Prof. dr. Jacob Seidell for valuable advice
during the study; the members of our advisory board: Sera de Vries, teacher, Edgar
van Mil, MD, PhD, pediatric–endocrinologist and Goof Buijs, M.Sc., senior advisor at
Netherlands Institute for Health Promotion and Disease Prevention for creative
suggestions; Herbert van den Heuvel and colleagues, for their dedication in the
beverage distribution; Ellinore Tellegen, for acting as independent physician.
64
DESIGN OF THE DRINK TRIAL
References
1. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence
of overweight and obesity in the United States, 1999-2004. JAMA 2006;295:1549-55.
2. Ogden CL, Carroll MD, Curtin LR, Lamb MM, Flegal KM. Prevalence of high body
mass index in US children and adolescents, 2007-2008. JAMA 2010;303:242-9.
3. Hayman LL, Meininger JC, Daniels SR, et al. Primary prevention of cardiovascular
disease in nursing practice: focus on children and youth: a scientific statement from
the American Heart Association Committee on Atherosclerosis, Hypertension, and
Obesity in Youth of the Council on Cardiovascular Disease in the Young, Council on
Cardiovascular Nursing, Council on Epidemiology and Prevention, and Council on
Nutrition, Physical Activity, and Metabolism. Circulation 2007;116:344-57.
4. Hannon TS, Rao G, Arslanian SA. Childhood obesity and type 2 diabetes mellitus.
Pediatrics 2005;116:473-80.
5. Steinberger J, Daniels SR. Obesity, insulin resistance, diabetes, and
cardiovascular risk in children: an American Heart Association scientific statement
from the Atherosclerosis, Hypertension, and Obesity in the Young Committee (Council
on Cardiovascular Disease in the Young) and the Diabetes Committee (Council on
Nutrition, Physical Activity, and Metabolism). Circulation 2003;107:1448-53.
6. Magarey AM, Daniels LA, Boulton TJ, Cockington RA. Predicting obesity in early
adulthood from childhood and parental obesity. Int J Obes Relat Metab Disord
2003;27:505-13.
7. Lobstein T, Baur L, Uauy R. Obesity in children and young people: a crisis in
public health. Obes Rev 2004;5 Suppl 1:4-82.
8. Franks PW, Hanson RL, Knowler WC, Sievers ML, Bennett PH, Looker HC.
Childhood obesity, other cardiovascular risk factors, and premature death. N Engl J
Med 2010;362:485-93.
9. Weiss R, Dziura J, Burgert TS, et al. Obesity and the metabolic syndrome in
children and adolescents. N Engl J Med 2004;350:2362-74.
10. Duffey KJ, Popkin BM. Shifts in patterns and consumption of beverages between
1965 and 2002. Obesity (Silver Spring) 2007;15:2739-47.
11. Nielsen SJ, Popkin BM. Changes in beverage intake between 1977 and 2001. Am
J Prev Med 2004;27:205-10.
12. Mattes R. Fluid calories and energy balance: the good, the bad, and the
uncertain. Physiol Behav 2006;89:66-70.
65
CHAPTER 3
13. Schulze MB, Manson JE, Ludwig DS, et al. Sugar-sweetened beverages, weight
gain, and incidence of type 2 diabetes in young and middle-aged women. JAMA
2004;292:927-34.
14. Palmer JR, Boggs DA, Krishnan S, Hu FB, Singer M, Rosenberg L. Sugar-sweetened
beverages and incidence of type 2 diabetes mellitus in African American women. Arch
Intern Med 2008;168:1487-92.
15. Bes-Rastrollo M, Sanchez-Villegas A, Gomez-Gracia E, Martinez JA, Pajares RM,
Martinez-Gonzalez MA. Predictors of weight gain in a Mediterranean cohort: the
Seguimiento Universidad de Navarra Study 1. Am J Clin Nutr 2006;83:362-70; quiz 94-5.
16. Nicklas TA, Yang SJ, Baranowski T, Zakeri I, Berenson G. Eating patterns and
obesity in children. The Bogalusa Heart Study. Am J Prev Med 2003;25:9-16.
17. Ludwig DS, Peterson KE, Gortmaker SL. Relation between consumption of sugarsweetened drinks and childhood obesity: a prospective, observational analysis. Lancet
2001;357:505-8.
18. Dubois L, Farmer A, Girard M, Peterson K. Regular sugar-sweetened beverage
consumption between meals increases risk of overweight among preschool-aged
children. J Am Diet Assoc 2007;107:924-34.
19. Johnson L, Mander AP, Jones LR, Emmett PM, Jebb SA. Is sugar-sweetened
beverage consumption associated with increased fatness in children? Nutrition
2007;23:557-63.
20. Newby PK, Peterson KE, Berkey CS, Leppert J, Willett WC, Colditz GA. Beverage
consumption is not associated with changes in weight and body mass index among
low-income preschool children in North Dakota. J Am Diet Assoc 2004;104:1086-94.
21. O'Connor TM, Yang SJ, Nicklas TA. Beverage intake among preschool children
and its effect on weight status. Pediatrics 2006;118:e1010-8.
22. Vandenbroucke JP. Observational research, randomised trials, and two views of
medical science. PLoS Med 2008;5:339-43.
23. Fewell Z, Davey Smith G, Sterne JAC. The Impact of Residual and Unmeasured
Confounding in Epidemiologic Studies: A Simulation Study. Am J Epidemiol
2007;166:646-55.
24. DiMeglio DP, Mattes RD. Liquid versus solid carbohydrate: effects on food intake
and body weight. Int J Obes Relat Metab Disord 2000;24:794-800.
25. Tordoff MG, Alleva AM. Effect of drinking soda sweetened with aspartame or
high-fructose corn syrup on food intake and body weight. Am J Clin Nutr 1990;51:963-9.
66
DESIGN OF THE DRINK TRIAL
26. Raben A, Vasilaras TH, Moller AC, Astrup A. Sucrose compared with artificial
sweeteners: different effects on ad libitum food intake and body weight after 10 wk
of supplementation in overweight subjects. Am J Clin Nutr 2002;76:721-9.
27. Chen L, Appel LJ, Loria C, et al. Reduction in consumption of sugar-sweetened
beverages is associated with weight loss: the PREMIER trial. Am J Clin Nutr
2009;89:1299-306.
28. James J, Thomas P, Cavan D, Kerr D. Preventing childhood obesity by reducing
consumption of carbonated drinks: cluster randomised controlled trial. BMJ
2004;328:1237-43.
29. Ebbeling CB, Feldman HA, Osganian SK, Chomitz VR, Ellenbogen SJ, Ludwig DS.
Effects of decreasing sugar-sweetened beverage consumption on body weight in
adolescents: a randomized, controlled pilot study. Pediatrics 2006;117:673-80.
30. Sichieri R, Paula Trotte A, de Souza RA, Veiga GV. School randomised trial on
prevention of excessive weight gain by discouraging students from drinking sodas.
Public Health Nutr 2009;12:197-202.
31. Katan MB. Weight-loss diets for the prevention and treatment of obesity. N Engl
J Med 2009;360:923-5.
32. Crawford PB, Woodward-Lopez G, Ritchie L, Webb K. How discretionary can we
be with sweetened beverages for children? J Am Diet Assoc 2008;108:1440-4.
33. Allison DB, Mattes RD. Nutritively sweetened beverage consumption and
obesity: the need for solid evidence on a fluid issue. JAMA 2009;301:318-20.
34. Scientific Committee on Food. Sucralose (opinion adapted by the SCF on 7
September 2000). http://ec.europa.eu/food/fs/sc/scf/out68_en.pdf. Accessed at
February 19, 2011.
35. Scientific Committee on Food. Re-evaluation of acesulfame K with reference to
the previous SCF opinion of 1991 (opinion expressed on 9th March 2000).
http://ec.europa.eu/food/fs/sc/scf/out52_en.pdf.). Accessed at February 19, 2011.
36. Snedecor GW, Cochran WG. Statistical Methods. 8th ed. Ames, USA: Iowa State
University Press; 1989.
37. Katan MB, Ludwig DS. Extra calories cause weight gain--but how much? JAMA
2010;303:65-6.
38. Wang YC, Gortmaker SL, Sobol AM, Kuntz KM. Estimating the energy gap among
US children: a counterfactual approach. Pediatrics 2006;118:e1721-33.
39. Federal Register. National Bureau of Standards. Customary system of weight and
measures. Federal Register 1968;33:34-5.
67
CHAPTER 3
40. Fredriks AM, van Buuren S, Burgmeijer RJ, et al. Continuing positive secular
growth change in The Netherlands 1955-1997. Pediatr Res 2000;47:316-23.
41. de Jong E, Schokker DF, Visscher TL, Seidell JC, Renders CM. Behavioural and
socio-demographic characteristics of Dutch neighbourhoods with high prevalence of
childhood obesity. Int J Pediatr Obes 2011;6:298-305.
42. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for
child overweight and obesity worldwide: international survey. BMJ 2000;320:1240-3.
43. Cole TJ, Flegal KM, Nicholls D, Jackson AA. Body mass index cut offs to define
thinness in children and adolescents: international survey. BMJ 2007;28:166-7.
44. CBS Statline Netherlands. Number of non-western people in the Netherlands.
http://statline.cbs.nl. Accessed at February 19, 2011.
45. CBS Statline Netherlands. Level of education of the Netherlands population (1565 year olds). http://statline.cbs.nl. Accessed at February 19, 2011.
46. Schulz KF, Grimes DA. Generation of allocation sequences in randomised trials:
chance, not choice. Lancet 2002;359:515-9.
47. Gibson RS. Principles of Nutritional Assessment. 2nd ed. New York, USA: Oxford
University Press; 2005.
48. Heymsfield SB, Lohman TG, Wang Z, Going SB. Human Body Composition 2nd ed.
Champaign, USA: Human Kinetics; 2005.
49. Roberts A, Renwick AG, Sims J, Snodin DJ. Sucralose metabolism and
pharmacokinetics in man. Food Chem Toxicol 2000;38 Suppl 2:S31-41.
50. Garnett SP, Baur LA, Cowell CT. Waist-to-height ratio: a simple option for
determining excess central adiposity in young people. Int J Obes (Lond) 2008;32:1028-30.
51. Deurenberg P, Pieters JJ, Hautvast JG. The assessment of the body fat
percentage by skinfold thickness measurements in childhood and young adolescence.
Br J Nutr 1990;63:293-303.
52. Rush EC, Puniani K, Valencia ME, Davies PS, Plank LD. Estimation of body fatness
from body mass index and bioelectrical impedance: comparison of New Zealand
European, Maori and Pacific Island children. Eur J Clin Nutr 2003;57:1394-401.
53. Rosenman K. Benefits of saccharin: a review. Environ Res 1978;15:70-81.
54. Davidson TL, Swithers SE. A Pavlovian approach to the problem of obesity. Int J
Obes Relat Metab Disord 2004;28:933-5.
55. Swithers SE, Baker CR, Davidson TL. General and persistent effects of highintensity sweeteners on body weight gain and caloric compensation in rats. Behav
Neurosci 2009;123:772-80.
68
DESIGN OF THE DRINK TRIAL
56. Swithers SE, Davidson TL. A role for sweet taste: calorie predictive relations in
energy regulation by rats. Behav Neurosci 2008;122:161-73.
57. Rodin J. Comparative effects of fructose, aspartame, glucose, and water preloads
on calorie and macronutrient intake. Am J Clin Nutr 1990;51:428-35.
58. Black RM, Tanaka P, Leiter LA, Anderson GH. Soft drinks with aspartame: effect
on subjective hunger, food selection, and food intake of young adult males. Physiol
Behav 1991;49:803-10.
59. Ford HE, Peters V, Martin NM, et al. Effects of oral ingestion of sucralose on gut
hormone response and appetite in healthy normal-weight subjects. Eur J Clin Nutr
2011;65:508-13.
69
CHAPTER 4
A trial of sugar-free or sugar-sweetened drinks and
body weight in children
Janne C de Ruyter
Margreet R Olthof
Jacob C Seidell
Martijn B Katan
N Engl J Med 2012;367:1397-1406
CHAPTER 4
Abstract
Background
The consumption of beverages that contain sugar is associated with overweight,
possibly because liquid sugars do not lead to a sense of satiety, so the consumption of
other foods is not reduced. However, data are lacking to show that the replacement
of sugar-containing beverages with noncaloric beverages diminishes weight gain.
Methods
We conducted an 18-month trial involving 641 primarily normal-weight children from
4 years 10 months to 11 years 11 months of age. Participants were randomly assigned
to receive 250 ml (8 oz) per day of a sugar-free, artificially sweetened beverage
(sugar-free group) or a similar sugar-containing beverage that provided 104 kcal
(sugar group). Beverages were distributed through schools. At 18 months, 26% of the
children had stopped consuming the beverages; the data from children who did not
complete the study were imputed.
Results
The z score for the body-mass index (BMI, the weight in kilograms divided by the
square of the height in meters) increased on average by 0.02 SD units in the sugarfree group and by 0.15 SD units in the sugar group; the 95% confidence interval (CI) of
the difference was −0.21 to −0.05. Weight increased by 6.35 kg in the sugar-free
group as compared with 7.37 kg in the sugar group (95% CI for the difference, −1.54
to −0.48). The skinfold-thickness measurements, waist-to-height ratio, and fat mass
also increased significantly less in the sugar-free group. Adverse events were minor.
When we combined measurements at 18 months in 136 children who had
discontinued the study with those in 477 children who completed the study, the BMI z
score increased by 0.06 SD units in the sugar-free group and by 0.12 SD units in the
sugar group (P = 0.06).
Conclusions
Masked replacement of sugar-containing beverages with noncaloric beverages
reduced weight gain and fat accumulation in normal-weight children. (Funded by the
Netherlands Organization for Health Research and Development and others; DRINK
ClinicalTrials.gov number, NCT00893529.)
72
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
Introduction
The increased prevalence of obesity in children, a major health problem,1,2 has
coincided with a large increase in the consumption of sugar-sweetened beverages.3
These beverages are considered to be more fattening than solid foods because they
do not lead to a sense of satiety.4 Thus, children who increase their consumption of
sugar-sweetened beverages may not reduce their intake of calories from other foods
and beverages, with a resultant increase in total energy intake and weight gain.
Consumption of sugar-sweetened beverages has been associated with weight
gain in most observational studies,4-8 though not all such studies.9-10 However,
children who drink more sugar-sweetened beverages also tend to eat more fast food
and to watch more television.11 Most studies adjust statistically for such confounders,
but residual and unmeasured confounding cannot be ruled out.12-13
Results of available trials are inconclusive,14-19 possibly owing to small samples, a
lack of adequate placebos, a short duration of the study, lack of individual
randomization, or a combination of these factors. In addition, effectively blinded trials
are important, because when investigators and participants know which treatment
should cause weight loss, results may be biased.20
We conducted the Double-blind, Randomized Intervention Study in Kids
(DRINK)20 to examine the effect on weight gain of masked replacement of sugarsweetened beverages with noncaloric, artificially sweetened beverages. The doubleblind design permitted the study of physiological mechanisms that were independent
of behavioral cues and voluntary changes in consumption.
Methods
Design and study population
The study was an 18-month, double-blind, randomized, controlled trial involving
schoolchildren living in the community who were 4 years 10 months to 11 years 11
months of age. The design of the study has been described previously.20 The study
started on November 14, 2009, and ended on July 22, 2011. We recruited children at
eight elementary schools in an urban area near Amsterdam. Baseline characteristics,
including usual beverage consumption, were determined with the use of a
questionnaire. Children were eligible only if they commonly drank sugar-sweetened
beverages, because we considered it unethical to provide sugary beverages to
children who did not habitually consume such beverages. We excluded children with
73
CHAPTER 4
various medical conditions (see Table S4.1 in the Supplementary Appendix, available
with the full text of this article at NEJM.org). We enrolled and individually randomly
assigned 641 children, stratified according to school, sex, age, and initial body-mass
index (BMI, the weight in kilograms divided by the square of the height in meters)
(Fig. 4.1).20 Children in the same household received the same type of beverage, but
they were unaware of this assignment.
For each child enrolled in the study, written informed consent was provided by a
parent or guardian who had obtained assent from the child. The study protocol was
approved by the medical ethics committee of VU University Medical Center
Amsterdam and is available at NEJM.org. Refresco Benelux, the manufacturer of the
beverages, had no role in the design of the study, the accrual or analysis of the data,
or the preparation of the manuscript. All authors vouch for the accuracy of the data
and the fidelity of the study to the protocol.
Intervention
We provided children with 1 can per day of a noncaloric, artificially sweetened,
noncarbonated beverage or a sugar-containing noncarbonated beverage. We
developed custom drinks for this study to ensure that the sugar-free and sugarcontaining drinks tasted and looked essentially the same.20 We hired Refresco Benelux
to produce these beverages. The identical-looking 250-ml (8-oz) cans provided either
0 or 26 g of sucrose (0 or 104 kcal per day). The sugar-free beverages contained 34 mg
of sucralose and 12 mg of acesulfame potassium per can.21
Participating children received a box at school each week labeled with their
name and containing 8 cans, 1 for each day of the week plus 1 extra to be used as a
spare in case a can was misplaced. The teachers checked to see whether the children
consumed their beverage during the morning break in class and reminded them to
take cans home for the weekend and any holidays.
We measured body weight, height, skinfold thickness (of the biceps, triceps, and
subscapular and suprailiac regions), waist circumference, and arm-to-leg electrical
impedance, and we collected urine samples at 0, 6, 12, and 18 months.20 For children
who stopped drinking the beverages before the study was completed, measurements
were obtained if the parent or guardian consented and the child assented. Data were
available at 18 months for 136 children who did not complete the study and for 477
children who did complete the study. Each child was evaluated by the same
investigator throughout the study. Two specially trained researchers measured the
waist circumference and the thickness of four skinfolds.
74
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
699 Children were screened
58 Were excluded
26 Were ineligible
32 Declined to participate
641 Underwent randomization
319 Were assigned to sugar-free group
322 Were assigned to sugar group
38 Discontinued study before 6-mo
measurement
24 No longer liked the beverage
7 Objected to the measurements
2 Declined to participate further
3 Had adverse events
2 Moved
28 Discontinued study between 6-mo
and 12-mo measurement
21 No longer liked the beverage
4 Objected to the measurements
1 Had adverse events
2 Moved
28 Discontinued study between 12-mo
and 18-mo measurement
23 No longer liked the beverage
1 Objected to the measurements
4 Had adverse events
38 Discontinued study before 6-mo
measurement
21 No longer liked the beverage
8 Objected to the measurements
8 Had adverse events
1 Moved
20 Discontinued study between 6-mo
and 12-mo measurement
14 No longer liked the beverage
2 Objected to the measurements
3 Had adverse events
1 Moved
12 Discontinued study between 12-mo
and 18-mo measurement
10 No longer liked the beverage
2 Had adverse events
319 (100%) Were included in full cohort,
with missing data imputed for children
who withdrew early
225 (71%) Were included in analysis of
children who completed study
322 (100%) Were included in full cohort,
with missing data imputed for children
who withdrew early
252 (78%) Were included in analysis of
children who completed study
Figure 4.1 Screening, randomization, and follow-up of the study participants. A total of 164 children
stopped consuming the study beverages. Measurements in 136 of these children (79 children in the sugar-free
group and 57 in the sugar group) were available at 18 months. Thus, measurements in 28 children who did not
complete the study (15 children in the sugar-free group and 13 in the sugar group) were not available at 18
20
months. We randomly assigned 641 children, not 642, as previously reported, since after unblinding, one child
20
whom we believed to have undergone randomization had not undergone randomization.
75
CHAPTER 4
Adherence
We provided frequent incentives for schools, teachers, parents, and children,
including tournaments, newsletters, birthday cards, and small gifts to encourage
adherence. We requested that parents report adverse events by contacting us
through the e-mail address or telephone number printed on each beverage can.20 We
visited the schools at least once a month to ensure that the study beverages were
delivered correctly to the classrooms. We calculated the adherence rate per child
during school days from the number of cans returned empty, half-filled, or full during
one randomly selected week each month (Table S4.2 in the Supplementary Appendix).
We measured the sucralose concentration in urine as an additional compliance
marker.22
Statistical analysis
Our predefined primary analysis involved the 477 children who completed the study
(i.e., the children who consumed the study beverages throughout the study).20 We
also used multiple imputation to impute the outcome values for the 164 children who
did not complete the study at 18 months. We created 30 multiple imputed-data sets
with five iterations, using the multivariate imputation by chained-equations algorithm
in R software, version 2.13. Variables included in the imputation model were age at
baseline, race or ethnic group, parents’ level of education, sex, compliance, study
group, and baseline and 18-month measurements — when available — of the
outcome being predicted.
The primary outcome20 was the BMI z score (expressed as the number of
standard deviations by which the BMI differed from the mean for a child’s age and sex
in the Netherlands).23 Pre-specified secondary outcomes were the waist-to height
ratio, the sum of the four skinfold-thickness measurements, and fat mass determined
by means of electrical impedance.20 Additional outcomes were weight, height, z score
for height,23 waist circumference, and weight change adjusted for height change.
Responses of the two study groups to the beverages were compared with two-sided ttests. We performed adjusted analyses with linear regression, using SPSS software.
Pre-specified adjustments for interdependency of outcomes in siblings, degree of
adherence, and baseline values20 had negligible effects on outcomes (Tables S4.3,
S4.4, and S4.5 in the Supplementary Appendix).
76
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
Results
Participants
Baseline characteristics were similar in the two study groups, except for a difference
in the parents’ level of education (Table 4.1). The mean BMI z score was 0.03, which
corresponded with the 51st percentile of Dutch children.23 The SD of 1.02 confirmed
that our sample was representative of Dutch children, for whom the SD equals 1.00 by
definition. At baseline, the participants consumed a mean (±SD) of 1.02±0.20 sugarsweetened beverages in the classroom during the 10 a.m. break, Monday through
Friday, and 1.50±1.40 sugar-sweetened beverages per day during weekends.
The net duration of the study was a mean of 541±8 days.20 The percentage of
participants who consumed the study beverages decreased from 100% at the
beginning of the study to 88% at 6 months, 81% at 12 months, and 74% at 18 months
(Fig. 4.1). The major reasons for discontinuing the study were dislike of the beverage
(accounting for 69% of the children who discontinued the study) and minor adverse
events (13%); weight gain accounted for 4% of children who discontinued the study
(four in the sugar group and two in the sugar-free group) (Fig. 4.1, and Table S4.6 in
the Supplementary Appendix).
Adherence and blinding
A total of 26% of the participants stopped consuming the beverages. These children
had a slightly higher BMI at baseline, and their parents had completed fewer years of
school (Table S4.7 in the Supplementary Appendix). This difference in educational
levels theoretically might have influenced the effect of the beverages on weight loss.
During the first 6 months of the study, however, weight loss was the same among
children who ultimately completed the study as among children who discontinued the
study after 6 months or more (Fig. S4.1E in the Supplementary Appendix). The
proportion of children who were aware of the type of beverage they were consuming
was similar among children who did and those who did not complete the study (Table
S4.8 in the Supplementary Appendix). Also, most children who did not complete the
study were lean, and few children dropped out because of concern about weight
(Tables S4.6 and S4.7 in the Supplementary Appendix). Most children who stopped
drinking the study beverages did so because they no longer liked the beverages.
Analyses in which missing values were imputed (Table 4.2) also suggested that results
for the full cohort would have been similar to those for the children who completed
the study.
77
CHAPTER 4
Table 4.1 Baseline characteristics of the study participants.a
Characteristic
Sugar-free
group (N = 319)
46
8.2±1.8
Sugar group
(N = 322)
47
8.2±1.8
Female sex (%)
Age (yr)
Ancestry (%)b
Dutch
80
76
Non-western
19
22
Highest level of education attained by parent or guardian (%)c
Elementary, vocational, or technical school
17
10
High-school diploma
33
28
College or university degree
49
61
Weight (kg)
30.04±8.93 30.33±8.82
Height
Measured (cm)
132.1±12.5 133.0±12.7
z score (SD units above or below mean in the Netherlands)d
−0.09±1.00
0.03±0.99
BMI
Calculated BMI
16.9±2.6
16.8±2.6
BMI z score (SD units above or below mean in the Netherlands)d
0.06±1.00
0.01±1.04
Low (%)e
1
1
Normal (%)
80
81
Overweight (%)
16
15
Obese (%)
3
3
Sum of thicknesses of four skinfolds (mm)f
36.4±17.7
35.6±17.9
Waist-to-height ratio (%)
44.6±4.0
44.2±4.0
Fat mass on electrical impedance
Measured (kg)g
5.8±3.8
5.7±3.7
As % of body weight
18
18
a
Plus-minus values are means ± SD. There were no significant between-group differences at
baseline except for the educational level of the parent or guardian (P = 0.02). P values were based
on chi-square tests and t-tests.
b
Data were available for 633 children; eight households did not provide this information. A child
was classified as Dutch if both parents were born in the Netherlands and as non-Western if one or
both parents were born in Suriname, Dutch Antilles, Turkey, Morocco, Russia, Egypt, or Vietnam.
c
Data were available for 632 children; nine households did not provide this information. We based
the educational level on that of the parent or guardian who had the higher level of education.
d
The z scores for BMI and height were calculated with the use of the data described by Schonbeck
et al.23
e
We used international cutoff points for low and normal BMI24 and for overweight and obesity.25
f
Data were available for 640 children; 1 child declined to be measured.
g
Data were available for 637 children; 4 children declined to be measured.
78
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
The 477 children who completed the study consumed 5.8 cans, or 83% of the assigned
7 cans per week, with no difference according to the type of beverage consumed and
no changes over time (Table S4.2 in the Supplementary Appendix). The mean level of
urinary sucralose was 6.7±4.7 mg per liter in the sugar-free group and 0.1±0.3 mg per
liter in the sugar group (Fig. 4.2), indicating adherence in the group of children who
drank the artificially sweetened beverages.
At 18 months, 609 children were asked which type of beverage they thought
they had received (Table S4.8 in the Supplementary Appendix). Among 474 children
who completed the study, 48% in the sugar-free group and 50% in the sugar group
answered that they did not know, 36% in the sugar-free group and 27% in the sugar
group answered “artificially sweetened,” and the remainder said “sugar-sweetened.”
The proportion of participants who correctly responded “artificially sweetened” was
21% (95% confidence interval [CI], 12 to 30) higher (47 more children) than expected
by chance, as estimated with the “blinding index” described by Bang et al.27 In the
sugar group, the proportion of children who were aware of the type of beverage they
had consumed was 3% (95% CI, −12 to 6) lower (7 fewer children) than expected.
Among 135 children who did not complete the study, the beverage was correctly
identified by 12% more, or 9 more children, than expected, in the sugar-free group,
and by 1 less child than expected in the sugar group.
BMI z score and other end points
In the full cohort of 641 children, the mean BMI z score increased by 0.02±0.41 SD
units in the sugar-free group and by 0.15±0.42 SD units in the sugar group (Table 4.2)
when missing values were imputed. The mean difference of 0.13 SD units was
significant. The sugar-free group gained significantly less body fat, as evidenced by
skinfold thickness, waist-to-height ratio, and electrical impedance. The mean weight
increased by 6.35±3.07 kg in the sugar-free group and by 7.37±3.35 kg in the sugar
group. The mean difference of 1.01 kg (2.2 lb) was significant. The mean difference in
weight gain decreased to 0.82 kg (P = 0.002) when adjusted for height change. The
BMI adjusted for age28 increased 0.36 less in the sugar-free group than in the sugar
group (P = 0.001). An alternative method for handling missing data — namely,
complete case analysis with covariate adjustment29 — yielded very similar results and
levels of significance (Table S4.9 in the Supplementary Appendix).
Similar results were also seen in the 477 children who consumed the study
beverages for the full 18 months (74% of the children enrolled) (Table 4.2 and Fig.
4.3). Children in the sugar-free group who completed the study gained 35% less body
79
80
0.08±0.99
39.6±20.4
43.7±4.0
6.77±4.71
17.22±8.44
36.39±10.41
142.34±12.48
−0.07±0.99
62.22±7.97
36.4±17.7
44.6±4.0
5.76±3.85
17.91±7.01
30.04±8.93
132.06±12.55
−0.09±1.00
58.85±7.44
18 months
0.06±1.00
0 months
Sugar-free group (N=319)
6.35±3.07
10.28±1.91
0.03±0.27
3.37±2.97
3.2±8.8
−0.9±2.0
1.01±2.62
−0.70±5.31
0.02±0.41
Change
30.33±8.82
133.02±12.71
0.03±0.99
58.69±7.05
35.6±17.9
44.2±4.0
5.70±3.68
17.67±6.92
0.01±1.04
0 months
37.69±11.05
143.67±13.05
0.09±0.99
62.72±7.92
41.1±21.1
43.7±4.0
7.28±4.89
18.05±8.25
0.15±1.06
18 months
Sugar group (N=322)
7.37±3.35
10.65±1.97
0.06±0.27
4.03±3.12
5.5±10.2
−0.5±2.0
1.58±2.47
0.38±4.86
0.15±0.42
Change
−1.01 (−1.54 to −0.48)
−0.37 (−0.72 to −0.02)
−0.04 (−0.10 to 0.02)
−0.66 (−1.23 to −0.09
−2.2 (−4.0 to −0.4)
−0.4 (−1.0 to −0.0)
−0.57 (−1.02 to −0.12)
−1.07 (−1.99 to −0.15)
−0.13 (−0.21 to −0.05)
Difference in change
from baseline (95% CI)
< 0.001
0.04
0.17
0.02
0.02
0.05
0.02
0.02
0.001
P value for
b
difference
Primary end point:
c
BMI z score
0.05±0.99
0.07 ±0.98
0.02 ±0.40
−0.02 ±1.00
0.14 ±1.06
0.15 ±0.42
−0.13 (−0.20 to −0.06)
0.001
Secondary end points
Sum of thicknesses of four skinfolds (mm)
36.0 ±16.9
39.1 ±20.2
3.2 ±8.1
34.4 ±15.8
40.1±20.4
5.7 ±10.0
−2.5 (−4.2 to −0.8)
0.003
Waist-to-height ratio (%)
44.6 ±3.7
43.7 ±3.8
−0.9 ±1.9
44.1 ±3.7
43.7 ±4.1
−0.5 ±2.1
−0.4 (−0.8 to −0.1)
0.02
d
Fat mass on electrical impedance (kg)
5.61 ±3.44
6.65 ±4.29
1.02 ±1.68
5.45 ±3.21
7.02 ±4.40
1.57±2.05
−0.55 (−0.89 to −0.21)
0.001
d
Fat mass on electrical impedance (%)
17.69±6.63
17.11±7.22
−0.61±3.71
17.35±6.39
17.85±7.36
0.45±3.81
−1.05 (−1.73 to −0.37)
0.003
Other end points
Weight (kg)
29.76±8.44
36.09±10.19
6.33±2.71
29.75±8.20
37.06±10.66
7.30±3.39
−0.97 (−1.52 to −0.42)
0.001
Height (cm)
131.72±12.44
141.92±12.23
10.21±1.85
132.40±12.57
142.98±12.89
10.57±1.93
−0.36 (−0.71 to −0.02)
0.04
c
Height z score
−0.12±0.92
−0.12±0.95
−0.001±0.38
−0.002±1.01
0.06±0.99
0.06±0.44
−0.07 (−0.14 to 0.01)
0.08
Waist circumference (cm)
58.63±6.90
61.99±7.79
3.36±2.69
58.30±6.43
62.35±7.55
4.05±3.10
−0.69 (−1.22 to −0.17)
0.01
a
Values are means (±SD) or means with 95% confidence intervals. Children were considered to have completed the study if they consumed the beverages for the full 18 months. Mean changes for the
children who completed the study may differ slightly from the difference between means at 18 and 0 months because measurements were not available from a few participants at either time point. We
used R software, version 2.1, to impute end points for the 164 participants who discontinued the study and SPSS software, version 17.0, for all other analyses.
b
Differences in changes from baseline between the sugar-free group and the sugar group were analyzed with the use of an independent-sample t-test; P≤0.05 was considered to indicate significance.
c
23
The z scores for body-mass index and height were calculated with the use of the data described by Schonbeck et al.
d
26
Impedance measurements were calculated with the use of the method described by Rush et al.
e
A total of 225 children in the sugar-free group and 252 children in the sugar group completed the study.
Primary end point
c
BMI z score
Secondary end points
Sum of thicknesses of four skinfolds (mm)
Waist-to-height ratio (%)
d
Fat mass on electrical impedance (kg)
d
Fat mass on electrical impedance (%)
Other end points
Weight (kg)
Height (cm)
c
Height z score
Waist circumference (cm)
e
Children who completed the study
Full cohort, with imputed data
Outcome
Table 4.2 Primary and secondary outcomes in the full cohort, with imputed data for children who did not complete the study, and in the
cohort of children who completed the study.a
CHAPTER 4
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
Figure 4.2 Urinary sucralose concentrations. The sucralose concentration was determined
21
in spot urine samples by means of liquid chromatography with mass spectrometry.
Samples were obtained from randomly selected children who completed the study. We
assigned a value of 0.01 to samples below the detection limit of 0.02 mg per liter. The
th
upper and lower ends of the boxes indicate the 25th and 75 quartiles, the black dots
means, the horizontal lines within the boxes medians, the upper whisker the maximum
value, and the lower whisker the minimum value. Values for the sugar-free group are
based on samples obtained from 116 children at 6 months and from 117 children at 12
and 18 months. Mean ±SD urinary sucralose concentrations were 6.3±3.7 mg per liter at 6
months, 6.6±4.5 mg per liter at 12 months, and 7.0±5.6 mg per liter at 18 months;
sucralose was undetectable in 3% of samples at 6 months, 8% of samples at 12 months,
and 10% of samples at 18 months. Values for the sugar group are based on samples
obtained from 54 children at 6 months and 36 children at 12 and 18 months. Mean values
were 0.04±0.13 mg per liter at 6 months, 0.03±0.14 mg per liter at 12 months, and
0.31±0.56 mg per liter at 18 months; sucralose was undetectable in 93% of samples at 6
months, 97% of samples at 12 months, and 67% of samples at 18 months. We also pooled
543 samples from participants at baseline to produce 20 pools. The mean sucralose
concentration in these samples was 0.06±0.07 mg per liter.
81
CHAPTER 4
fat than those in the sugar group, according to impedance measurements,26 and 19%
less when fat mass was calculated from the sum of the thicknesses of four skinfolds.
According to the changes in skinfold thickness, the sugar-free group gained 1.47 kg of
body fat and the sugar group gained 1.82 kg.30 Most of the effect on BMI z score and
weight was achieved in the first 6 months (Fig. 4.3B, and Fig. S4.2B in the
Supplementary Appendix). The mean height increased by 10.21±1.85 cm in the sugarfree group and by 10.57±1.93 cm in the sugar group. The mean difference of 0.36 cm
(0.14 in) was significant (P = 0.04), but the difference in z score for height was not
significant (Table 4.2).
We obtained measurements at 18 months in 136 of the 164 children who did
not complete the study. When we combined their measurements with those in the
477 children who completed the study, the mean BMI z score increased by 0.06±0.44
SD units in the sugar-free group and by 0.12±0.44 SD units in the sugar group. The
mean difference of 0.07 SD units was not significant (95% CI, −0.134 to 0.002; P =
0.06).
Discussion
We found that masked replacement of a sugar-containing beverage with a sugar-free
beverage significantly reduced weight gain and body fat gain in healthy children. Our
study had several strengths. The double-blind design eliminated the effects of
psychological cues and socially desirable behavior and allowed testing of biologic
mechanisms alone. Although blinding was imperfect, it was more successful than in
most randomized, double-blind trials.31,32 Measurements of urinary sucralose levels
suggested a high rate of adherence. Previous trials may have yielded inconsistent
results because of small samples, short duration, poor adherence, or lack of individual
randomization.14-19 Our sample size was adequate to allow precise outcomes, and the
18-month study duration ensured that the observed effect was not transient. The
large sample and stratified randomization produced well-balanced study groups at
baseline. We assume that the mean changes in other factors that affect weight were
also similar between the groups. Thus, the observed differences in body fat and BMI z
score can be ascribed primarily to the assigned beverage.
Our study had certain limitations. A total of 26% of the participants did not
complete the study. However, as long as they were participating in the study, their
changes in weight and body fat paralleled those in children who ultimately completed
82
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
BMI z scores as function of time
0.17
Standard deviation units
0.15
0.13
0.11
0.09
0.07
0.05
0.03
0.01
-0.01
-0.03
0
A
6
12
18
Follow-up (months)
Sugar-free group
Sugar-containing group
Standard deviation units
Between-group difference in change
in BMI z score
0.02
0.00
-0.02
-0.04
-0.06
-0.08
-0.10
-0.12
-0.14
-0.16
0
6
12
18
Follow-up (months)
B
Figure 4.3 Body-Mass Index (BMI) z score in the 477 children who drank the study
beverages for the full 18 months. The z score for BMI is the BMI expressed as the number
of standard deviations by which a child differed from the mean in the Netherlands for his
or her age and sex. Panel A shows mean z scores for the two study groups over the 18month study period. Panel B shows the between-group difference in the mean change
from baseline (the mean change in the BMI z score in the sugar-free group minus the
mean change in the sugar group), as a function of time. T bars in both panels indicate
standard errors.
83
CHAPTER 4
the study. We therefore suggest that the study beverage was not inherently
inefficacious in these children. Data on weight and height at the end of the study were
available for most of the children. When we pooled those measurements with the
measurements for the children who completed the study, the effect of the study
beverage became smaller and non-significant. This finding was expected; the children
who did not complete the study probably went back to drinking sugary beverages,
which attenuated the effect of any sugar-free beverages that they consumed before
discontinuation.
Approximately 0.5 kg of the difference in weight gain between the two study
groups was due to fat mass. We speculate that another 0.3 kg may have been due to
the changes in lean mass that accompany changes in body fat.33-35 Thus, about 0.8 kg
of the difference in weight gain was probably due to body fat and associated muscle
and other tissues.33 Another 0.2 kg can be ascribed to the difference in height gain.23
Therefore, the increases in body weight as predicted from increases in fat mass and
height differed by about 1 kg between the study groups. This estimate is consistent
with the actual measured difference in body-weight change.
Although the difference in height gain was minute, it warrants scrutiny. Some
studies suggest that obese pre-pubertal children are indeed taller than normal-weight
children.36 However, obesity is associated with an earlier onset of puberty,37-39 which
predicts shorter stature in adults.40 The increase in BMI in Dutch children in the past
decades has not led to an increase in final height.23 We speculate that a modest
reduction of liquid calories in children will have little effect on adult height.
A plausible explanation for the observed reduction in body fat is that the
removal of liquid sugar was not sensed by satiating mechanisms and was incompletely
compensated for by the increased consumption of other foods.41 We speculate that
reduced ingestion of liquid sugars might also reduce the insulin spike and thus
diminish hunger.42 We find it less likely that our results were caused by the artificial
sweeteners in the sugar-free beverages, because sweeteners do not suppress caloric
intake.43,44 Therefore, we assume that water or other non-caloric beverages would be
as effective.
In observational studies, the consumption of artificially sweetened beverages is
associated with weight gain rather than weight loss.45,46 This finding has led to the
hypothesis that artificial sweeteners induce weight gain (e.g., by activating sweettaste receptors in the gut).47 Our findings do not support this hypothesis.
Alternatively, people who are at risk for gaining weight may turn to artificial
sweeteners in an attempt to reduce caloric intake.45,46 Consumers may also believe
84
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
that the use of such sweeteners permits them to eat more of other foods, but this
may lead to a net increase in total caloric intake.46 Whatever the explanation, the
epidemiologic association of the use of artificial sweeteners with obesity does show
that switching to artificially sweetened beverages by itself is insufficient to combat
weight gain.
The participants in our study were healthy Dutch children, most of whom were
white and of normal weight. Thus, we do not know whether the results would be
similar in other ethnic groups, obese children, or adults, though we speculate that the
same biologic mechanisms are operative. The findings of Ebbeling et al.,48 reported
elsewhere in this issue of the Journal, would support such speculation. The findings of
Qi et al.,49 reported elsewhere in this issue of the Journal, suggest that persons with a
genetic predisposition to obesity are especially susceptible to the effects of sugarsweetened beverages on BMI. Children in the United States consume on average
almost three times as many calories from sugar-sweetened beverages as the amount
provided in our trial.50 We speculate that decreased consumption of such beverages
might reduce the high prevalence of overweight in these children.
Supported by
Supported by grants from the Netherlands Organization for Health Research and
Development (120520010), the Netherlands Heart Foundation (2008B096), and the
Royal Netherlands Academy of Arts and Sciences (ISK/741/PAH). Disclosure forms
provided by the authors are available with the full text of this article at NEJM.org.
Acknowledgements
We thank the children and parents for their dedication; the school administrators and
teachers for their cooperation; Emilie de Zoete and Hetty Geerars for assistance in the
execution of this study; Joop Bremer for advice on development and manufacture of
beverages; Lothar Kuijper for statistical advice; Rolf Groenwold and Martijn Heijmans
for assistance with imputation analyses; our students for assistance in the data
collection; Herbert van den Heuvel and colleagues for beverage distribution; Sera de
Vries and colleagues for help with the pilot study; and Rob van Dam, Judith Neter, and
Tommy Visscher, for contributions during the preparation of the trial.
85
CHAPTER 4
References
1. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of obesity and trends in
body mass index among US children and adolescents, 1999-2010. JAMA 2012;307:
483-90.
2.
Olds T, Maher C, Zumin S, et al. Evidence that the prevalence of childhood
overweight is plateauing: data from nine countries. Int J Pediatr Obes 2011;6:342-60.
3. Duffey KJ, Popkin BM. Shifts in patterns and consumption of beverages between
1965 and 2002. Obesity (Silver Spring) 2007;15:2739-47.
4. Mattes R. Fluid calories and energy balance: the good, the bad, and the
uncertain. Physiol Behav 2006;89:66-70.
5. Dhingra R, Sullivan L, Jacques PF, et al. Soft drink consumption and risk of
developing cardiometabolic risk factors and the metabolic syndrome in middle aged
adults in the community. Circulation 2007;116:480-8.
6. Dubois L, Farmer A, Girard M, Peterson K. Regular sugar-sweetened beverage
consumption between meals increases risk of overweight among preschool-aged
children. J Am Diet Assoc 2007;107:924-34.
7. Ludwig DS, Peterson KE, Gortmaker SL. Relation between consumption of sugarsweetened drinks and childhood obesity: a prospective, observational analysis. Lancet
2001;357:505-8.
8.
Schulze MB, Manson JE, Ludwig DS, et al. Sugar-sweetened beverages, weight
gain, and incidence of type 2 diabetes in young and middle-aged women. JAMA
2004;292:927-34.
9. Newby PK, Peterson KE, Berkey CS, Leppert J, Willett WC, Colditz GA. Beverage
consumption is not associated with changes in weight and body mass index among
low-income preschool children in North Dakota. J Am Diet Assoc 2004;104: 1086-94.
10. O’Connor TM, Yang SJ, Nicklas TA. Beverage intake among preschool children
and its effect on weight status. Pediatrics 2006;118(4):e1010-e1018.
11. Park S, Blanck HM, Sherry B, Brener N, O’Toole T. Factors associated with sugar
sweetened beverage intake among United States high school students. J Nutr 2012;
142:306-12.
12. Fewell Z, Davey Smith G, Sterne JAC. The impact of residual and unmeasured
confounding in epidemiologic studies: a simulation study. Am J Epidemiol 2007;
166:646-55.
13. Vandenbroucke JP. Observational research, randomised trials, and two views of
medical science. PLoS Med 2008;5(3):e67.
86
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
14. Raben A, Vasilaras TH, Møller AC, Astrup A. Sucrose compared with artificial
sweeteners: different effects on ad libitum food intake and body weight after 10 wk
of supplementation in overweight subjects. Am J Clin Nutr 2002;76:721-9.
15. Tordoff MG, Alleva AM. Effect of drinking soda sweetened with aspartame or
high-fructose corn syrup on food intake and body weight. Am J Clin Nutr 1990;51:963-9.
16. Raben A, Møller BK, Flint A, et al. Increased postprandial glycaemia, insulinemia,
and lipidemia after 10 weeks’ sucrose-rich diet compared to an artificially sweetened
diet: a randomised controlled trial. Food Nutr Res 2011;55:5961-74.
17. Ebbeling CB, Feldman HA, Osganian SK, Chomitz VR, Ellenbogen SJ, Ludwig DS.
Effects of decreasing sugar-sweetened beverage consumption on body weight in
adolescents: a randomized, controlled pilot study. Pediatrics 2006;117:673-80.
18. James J, Thomas P, Cavan D, Kerr D. Preventing childhood obesity by reducing
consumption of carbonated drinks: cluster randomised controlled trial. BMJ 2004;
328:1237-43.
19. Tate DF, Turner-McGrievy G, Lyons E, 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;95:555-63.
20. de Ruyter JC, Olthof MR, Kuijper LD, Katan MB. Effect of sugar-sweetened
beverages on body weight in children: design and baseline characteristics of the
Double blind, Randomized INtervention study in Kids. Contemp Clin Trials
2012;33:247-57.
21. Yang DJ, Chen B. Simultaneous determination of nonnutritive sweeteners in
foods by HPLC/ESI-MS. J Agric Food Chem 2009;57:3022-7.
22. Roberts A, Renwick AG, Sims J, Snodin DJ. Sucralose metabolism and
pharmacokinetics in man. Food Chem Toxicol 2000;38:Suppl 2:S31-S41.
23. Schönbeck Y, Talma H, van Dommelen P, et al. Increase in prevalence of
overweight in Dutch children and adolescents: a comparison of nationwide growth
studies in 1980, 1997 and 2009. PLoS One 2011;6(11):e27608.
24. Cole TJ, Flegal KM, Nicholls D, Jackson AA. Body mass index cut offs to define
thinness in children and adolescents: international survey. BMJ 2007;225:194.
25. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for
child overweight and obesity worldwide: international survey. BMJ 2000; 3220:1240-3.
26. Rush EC, Puniani K, Valencia ME, Davies PS, Plank LD. Estimation of body fatness
from body mass index and bioelectrical impedance: comparison of New Zealand
European, Maori and Pacific Island children. Eur J Clin Nutr 2003;57:1394-401.
87
CHAPTER 4
27. Bang H, Ni L, Davis CE. Assessment of blinding in clinical trials. Control Clin Trials
2004;25:143-6.
28. Berkey CS, Colditz GA. Adiposity in adolescents: change in actual BMI works
better than change in BMI z score for longitudinal studies. Ann Epidemiol 2007; 17:44-50.
29. Groenwold RH, Donders AR, Roes KC, Harrell FE Jr, Moons KG. Dealing with
missing outcome data in randomized trials and observational studies. Am J Epidemiol
2012;175:210 -7.
30. Deurenberg P, Pieters JJ, Hautvast JG. The assessment of the body fat
percentage
by skinfold thickness measurements in childhood and young adolescence. Br J Nutr
1990;63:293-303.
31. Fergusson D, Glass KC, Waring D, Shapiro S. Turning a blind eye: the success of
blinding reported in a random sample of randomised, placebo controlled trials. BMJ
2004;328:432-4.
32. Hróbjartsson A, Forfang E, Haahr MT, Als-Nielsen B, Brorson S. Blinded trials
taken to the test: an analysis of randomized clinical trials that report tests for the
success of blinding. Int J Epidemiol 2007;36:654-63.
33. Ducher G, Bass SL, Naughton GA, Eser P, Telford RD, Daly RM. Overweight
children have a greater proportion of fat mass relative to muscle mass in the upper
limbs than in the lower limbs: implications for bone strength at the distal forearm. Am
J Clin Nutr 2009;90:1104-11.
34. Tremblay A, Després JP, Thériault G, Fournier G, Bouchard C. Overfeeding and
energy expenditure in humans. Am J Clin Nutr 1992;56:857-62.
35. Treuth MS, Butte NF, Sorkin JD. Predictors of body fat gain in nonobese girls with
a familial predisposition to obesity. Am J Clin Nutr 2003;78:1212-8.
36. Ajala O, Frémeaux AE, Hosking J, et al. The relationship of height and body fat to
gender-assortative weight gain in children: a longitudinal cohort study (Early- Bird 44).
Int J Pediatr Obes 2011;6:223-8.
37. Buyken AE, Karaolis-Danckert N, Remer T. Association of prepubertal body
composition in healthy girls and boys with the timing of early and late pubertal
markers. Am J Clin Nutr 2009;89:221-30.
38. Mumby HS, Elks CE, Li S, et al. Mendelian randomisation study of childhood BMI
and early menarche. J Obes 2011; 2011:180729.
39. Ong KK, Emmett P, Northstone K, et al. Infancy weight gain predicts childhood
body fat and age at menarche in girls. J Clin Endocrinol Metab 2009;94: 1527-32.
88
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
40. McIntyre MH. Adult stature, body proportions and age at menarche in the
United States National Health and Nutrition Survey (NHANES) III. Ann Hum Biol
2011;38:716-20. 41. Cassady BA, Considine RV, Mattes RD. Beverage consumption,
appetite, and energy intake: what did you expect? Am J Clin Nutr 2012;95:587-93.
42. Ludwig DS. The glycemic index: physiological mechanisms relating to obesity,
diabetes, and cardiovascular disease. JAMA 2002;287:2414-23.
43. Blundell JE, Green SM. Effect of sucrose and sweeteners on appetite and energy
intake. Int J Obes Relat Metab Disord 1996;20:Suppl 2:S12-S17.
44. Renwick AG, Molinary SV. Sweet-taste receptors, low-energy sweeteners,
glucose absorption and insulin release. Br J Nutr 2010;104:1415-20.
45. Foreyt J, Kleinman R, Brown R, Lindstrom R. The use of low-calorie sweeteners
by children: implications for weight management. J Nutr 2012;142:1155S-1162S.
46. Fowler SP, Williams K, Resendez RG, Hunt KJ, Hazuda HP, Stern MP. Fueling the
obesity epidemic? Artificially sweetened beverage use and long-term weight gain.
Obesity (Silver Spring) 2008;16:1894- 900.
47. Pepino MY, Bourne C. Non-nutritive sweeteners, energy balance, and glucose
homeostasis. Curr Opin Clin Nutr Metab Care 2011;14:391-5.
48. Ebbeling CB, Feldman HA, Chomitz VR, et al. A randomized trial of sugar
sweetened beverages and adolescent body weight. N Engl J Med;367:1407-16
49. Qi Q, Chu AY, Kang JH, et al. Sugar sweetened beverages and genetic risk of
obesity. N Engl J Med 2012;367:1387-96
50. Wang YC, Bleich SN, Gortmaker SL. Increasing caloric contribution from sugar
sweetened beverages and 100% fruit juices among US children and adolescents, 19882004. Pediatrics 2008;121(6):e1604-e1614.
89
CHAPTER 4
Supplementary Appendix
Figure S4.1
Figure S4.2
Figure S4.3
Table S4.1
Table S4.2
Table S4.3
Table S4.4
Table S4.5
Table S4.6
Table S4.7
Table S4.8
Table S4.9
90
Comparison of outcomes after 6 months in completers and future
noncompleters.
Time course of body weight in the 477 completers.
Time course of height in the 477 completers.
Inclusion and exclusion criteria for participation.
Adherence based on can counts.
Results considering siblings.
Results in the 80% most adherent participants.
Linear regression analyses to correct for baseline differences.
Adverse events.
Baseline characteristics of completers versus noncompleters.
Success of blinding.
Complete case analysis with covariate adjustment.
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
A
B
C
D
E
F
Figure S4.1 Comparison of outcomes after 6 months in completers and future noncompleters.
Diamonds indicate differences in response between treatment groups among the 88 children
who still consumed their drinks at 6 months but became noncompleters later; each point
represents the mean of 56 children in the sugar-free group minus the mean of 32 in the sugarcontaining group. Triangles indicate differences in response between treatment groups among
the completers, i.e. the 477 children who drank the beverages for the full 18 months; each
point represents the mean of 225 children in the sugar-free group minus the mean of 252 in
the sugar-containing group. Bars indicate one standard error.
91
CHAPTER 4
Difference between treatment
groups in the change of body weight
with time
Body weight as function of time
40
0.2
38
0.0
36
kg
kg
-0.2
34
-0.4
32
-0.6
30
-0.8
28 0
-1.0
0
6
12
18
0
Follow-up (months)
A
Sugar-containing group
6
12
18
Follow-up (months)
Sugar-free group
B
Figure S4.2 Time course of body weight in the 477 completers. Panel A shows the mean
body weight for the sugar-free and the sugar-containing group over the course of the trial.
Panel B shows the mean body weight gain in the sugar-free group minus that in the sugarcontaining group. Bars indicate one standard error.
Difference between treatment
groups in the change of height with
time
Height as function of time
0.1
146
144
0.0
142
cm
cm
140
138
136
-0.1
-0.2
134
132
-0.3
130
128
-0.4
0
A
6
12
18
Follow-up (months)
Sugar-containing group
Sugar-free group
0
6
12
18
Follow-up (months)
B
Figure S4.3 Time course of height in the 477 completers. Panel A shows mean heights for
the sugar-free and the sugar-containing group over the course of the study. Panel B shows
the height gain in the sugar-free group minus the height gain in the sugar-containing
group. Bars indicate one standard error.
92
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
Table S4.1 Inclusion and exclusion criteria for participation.
Inclusion criteria:
· Habitual consumption of 250 mL or more per day of sugar-containing beverages on at
a
least three out of five school days
· Minimum age of 5 years at the start of the trial
· Young enough to be still in elementary school at the end of the trial
· Written informed consent by a parent or guardian
Exclusion criteria:
· Medication or medical treatment for obesity
· Diabetes, growth disorders, celiac disease, or serious gastro-enterologic diseases (for
example inflammatory bowel disease).
· Medical history or surgery known to interfere with the study
· Participation in another intervention study up to 3 months before and during the trial
if the intervention interfered with our study
· Physical disabilities that hamper the measurements
· Intention to change location of residence and elementary school during the trial
a
At baseline the participants actually consumed 1.02±0.2 (mean±SD) sugar-sweetened
beverages in the classroom during the 10 am morning break, Monday through Friday, and
1.5±1.4 (mean±SD) sugar-sweetened beverages per day on Saturday and Sunday.
93
CHAPTER 4
a
Table S4.2 Adherence as estimated from can counts.
Treatment
Number of can
Number of
% of assigned cans
counts during
completers available returned empty,
trial (weeks)
per can count,
mean±SD
mean±SD
Completers
Sugar-free group
17
205 ±16
84.7 ±24.4
Sugar-containing group
17
228 ±18
85.2 ±24.3
a
We measured compliance by counting returned school day cans. Children were
instructed to drink one can per day at school Monday to Friday, and to place both empty
and full cans back into their kid boxes to be collected by us. We counted returned cans of
one week each month. Empty cans counted as 100% adherence, half -filled cans as 50%
and full cans as 0%. Lowest adherence was 74% in April 2010, 5 months into the study,
and highest adherence was 92%, in September 2010.
Cans emptied during weekends and holidays were not returned, but cans for weekends
and holidays that had been left uncollected at school were returned. In order to estimate
adherence during weekends, we checked how many cans the children took home. We
provided 2 cans for each weekend plus one spare can. Completers returned a mean±SD of
1.18±1.10 full cans per week. Thus they took on average 1.82 full cans home per weekend,
or 0.91 per weekend day. We assume that 15% of this was wasted and 85% consumed.
Total consumption per week would then be 85% of 1.82 cans per weekend plus 85% of 5
cans per school week, for a total of 5.8 cans. That equals 83% of the planned 7 cans.
94
0 month
18 months
Sugar-free group
N=199
Change
Difference in changes P for
(95% CI)
difference
Completers onlyb
Primary endpoint
BMI z scorec
0.07±0.92
0.09±0.92
0.02±0.40
−0.002±0.92
0.14±0.98
0.15±0.39 −0.13 (−0.21 to −0.05)
0.002
Secondary endpoints
Sum of skinfolds (mm)
36.5±16.0
39.6±19.2
3.1±8.2
34.5±14.2
40.2±18.8
5.7±9.3
−2.7 (−4.5 to −0.9)
0.004
Waist-to-height ratio (%)
44.6±3.5
43.7±3.7
−0.9±1.8
44.1±3.4
43.6±3.9
−0.46±2.01
−0.4 (−0.8 to −0.04)
0.03
Electrical-impedance fat mass (kg) 5.66±3.22
6.71±3.99
1.03±1.61
5.53±2.97
7.15±4.03
1.63±1.87 −0.60 (−0.95 to −0.24)
0.001
Electrical-impedance fat mass
17.97±6.41
17.43±6.82
−0.58±3.51
17.63±6.18
18.21±6.88
0.58±3.57 −1.16 ( −1.89 to −0.44)
0.002
(% of body weight)
Other endpoints
Weight (kg)
29.83±7.84
36.19±9.53
6.36±2.62
30.02±7.53
37.35±9.74
7.33±3.13 −0.97 (−1.56 to −0.38)
0.001
Height (cm)
131.76±11.49 141.98±11.38 10.22±1.74 132.94±11.47 143.49±11.70 10.55±1.83 −0.33 (−0.69 to 0.04)
0.08
Height z scorec
−0.13±0.90
−0.13±0.92
0.02±0.37
0.02±0.98
0.09±0.95
0.07±0.44
−0.05 (−1.14 to 0.03)
0.21
Waist circumference (cm)
58.73±6.53
62.09±7.43
3.36±2.60
58.50±5.93
62.52±6.97
4.02±3.00 −0.67 (−1.24 to −0.09)
0.02
a
This was a prespecified analysis.1 Values are means±SD, or means with 95% confidence intervals in parentheses. N is the number of independent households. 119
household provided more than one participant, and values of siblings were averaged to provide a single experimental unit. We used the Statistical Package for the
Social Sciences (SPSS) version 17.0 to perform the analyses. Differences in change between the sugar-free group en sugar-containing group were analysed with an
independent sample t-test: we considered P ≤ 0.05 significant.
b
Completers were participants who consumed the drinks for the full 18 months. Mean changes may differ slightly from the difference between means at 18 and 0
months because a few participants lacked measurements at either timepoint.
c
We calculated z scores of body-mass index and height from the Dutch 2009 reference data.2
0 month
18 months
Change
Sugar-containing group
N=176
Table S4.3 Results when data of siblings were collapsed into a single experimental unit.a
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
95
96
18 months
Change
0 months
18 months
Change
Difference in changes
(95% CI)
P for
difference
Sugar-free group
Sugar-containing group
Completers only b
N=174
N=209
Primary endpoint
BMI z score c
0.04±0.97
0.06±0.96 0.02±0.41 0.01±1.01
0.17±1.05 0.16±0.40
−0.14 (−0.23 to −0.06)
0.001
Secondary endpoints
Sum of skinfolds (mm)
35.3±16.9
38.2±19.9
2.9±8.2
34.1±15.7
39.7±20.0 5.7±10.2
−2.7 (−4.6 to −0.8)
0.005
Waist-to-height ratio (%)
44.6±3.7
43.6±3.8
−1.0±1.9
44.2±3.6
43.8±4.0
−0.4±2.1
−0.5 (−0.9 to −0.1)
0.01
Electrical-impedance fat mass (kg)
5.49±3.44
6.44±4.17 0.95±1.60 5.35±3.09
6.88±4.20 1.53±1.98
−0.59 (−0.95 to −0.22)
0.002
Electrical-impedance fat mass
17.39±6.60 16.70±7.00 −0.68±3.76 17.22±6.22 17.78±7.27 0.50±3.82
−1.18 (−1.95 to −0.41
0.003
(% of body weight)
Other endpoints
Weight (kg)
29.59±8.47 35.84±10.08 6.25±2.63 29.49±8.17 36.64±10.38 7.14±3.18
−0.89 (−1.47 to −0.30)
0.003
Height (cm)
131.54±12.20 141.84±12.00 10.30±1.91 131.83±12.27 142.32±12.56 10.48±1.92
−0.18 (−0.57 to 0.20)
0.35
Height z scorec
−0.06±0.91 −0.06±0.95 −0.005±0.42 0.05±1.00
0.09±0.98 0.05±0.46
−0.05 (−0.14 to 0.04)
0.26
Waist circumference (cm)
58.57±7.01 61.86±7.67 3.29±2.62 58.15±6.42 62.20±7.32 4.05±3.08
−0.76 (−1.34 to −0.18)
0.01
a
This was a prespecified analysis.1 Values are means±SD, or means with 95% confidence intervals in parentheses. We excluded children with an adherence of
77.3% or less. We used the Statistical Package for the Social Sciences (SPSS) version 17.0 to perform the analyses. Differences in change between the sugar-free
group en sugar-containing group were analyzed with an independent sample t-test: we considered P ≤ 0.05 significant.
b
Completers were participants who consumed the drinks for the full 18 months. Mean changes may differ slightly from the difference between means at 18
and 0 months because a few participants lacked measurements at either timepoint.
c
We calculated z scores of body-mass index and height from the Dutch 2009 reference data.2
0 months
Table S4.4 Results in the 80% most compliant participants.a
CHAPTER 4
39.6±20.4
43.7±4.0
6.77±4.71
17.22±8.44
36.39±10.41
142.34±12.48
−0.07±0.99
62.22±7.97
36.4±17.7
44.6±4.0
5.76±3.85
17.91±7.01
30.04±8.93
132.06±12.55
−0.09±1.00
58.85±7.44
N=225
0.08±0.99
0.06±1.00
N=319
Sugar-free group
18 months
6.35±3.07
10.28±1.91
0.03±0.27
3.37±2.97
3.2±8.8
−0.9±2.0
1.01±2.62
−0.70±5.31
0.02±0.41
Change
30.33±8.82
133.02±12.71
0.03±0.99
58.69±7.05
35.6±17.9
44.2±4.0
5.70±3.68
17.67±6.92
0.01±1.04
0 months
N=252
37.69±11.05
143.67±13.05
0.09±0.99
62.72±7.92
41.1±21.1
43.7±4.0
7.28±4.89
18.05±8.25
0.15±1.06
N=322
Sugar-containing group
18 months
7.37±3.35
10.65±1.97
0.06±0.27
4.03±3.12
5.5±10.2
−0.5±2.0
1.58±2.47
0.38±4.86
0.15±0.42
Change
−0.97 (−1.46 to −0.48)
−0.37 (−0.72 to −0.02)
−0.04(−0.10 to 0.02)
−0.66 (−1.23 to −0.09)
−2.3 (−4.1 to −0.5)
−0.004 (−0.01 to −0.0)
−0.57 (−1.02 to −0.12)
−1.06 (−1.98 to −0.14)
−0.12 (−0.20 to −0.04)
Difference in changes
(95% CI)
< 0.001
0.05
0.12
0.02
0.01
0.08
0.01
0.02
0.001
P for
difference
Primary endpoint
BMI z scored
0.05±0.99
0.07 ±0.98
0.02 ±0.40
−0.02 ±1.00
0.14 ±1.06
0.15 ±0.42
−0.13 (−0.20. to −0.05)
0.001
Secondary endpoints
Sum of four skinfolds (mm)
36.0 ±16.9
39.1 ±20.2
3.2 ±8.1
34.4 ±15.8
40.1±20.4
5.7 ±10.0
−2.7 (−4.3 to −1.05)
0.001
Waist-to-height ratio (%)
44.6 ±3.7
43.7 ±3.8
−0.9 ±1.9
44.1 ±3.7
43.7 ±4.1
−0.5 ±2.1
−0.4 (−0.8 to −0.1)
0.03
Electrical-impedance fat mass (kg)
5.61 ±3.44
6.65 ±4.29
1.02 ±1.68
5.45 ±3.21
7.02 ±4.40
1.57±2.05
−0.58 (−0.90 to −0.26)
< 0.001
Electrical-impedance fat mass (% of body weight)
17.69±6.63
17.11±7.22
−0.61±3.71
17.35±6.39
17.85±7.36
0.45±3.81
−1.04 (−1.72 to −0.36)
0.003
Other endpoints
Weight (kg)
29.76±8.44
36.09±10.19
6.33±2.71
29.75±8.20
37.06±10.66
7.30±3.39
−0.97 (−1.42 to −0.52)
< 0.001
Height (cm)
131.72±12.44
141.92±12.23
10.21±1.85
132.40±12.57
142.98±12.89
10.57±1.93
−0.37 (−0.71 to −0.03)
0.04
d
Height z score
−0.12±0.92
−0.12±0.95
−0.001±0.38
−0.002±1.01
0.06±0.99
0.06±0.44
−0.08 (−0.15 to −0.003)
0.04
Waist circumference (cm)
58.63±6.90
61.99±7.79
3.36±2.69
58.30±6.43
62.35±7.55
4.05±3.10
−0.70 (−1.24 to −0.20)
0.007
a
This was a prespecified analysis where we used the difference in change between the sugar-containing and sugar-free group as dependent variable, treatment as independent variable, and the baseline
value of each endpoint as covariate.1 Values are means±SD, or means with 95% confidence intervals in parentheses. We used R software version 2.1 to impute endpoints of 164 participants who discontinued
treatment, and the Statistical Package for the Social Sciences (SPSS) version 17.0 for all other analyses. The difference in changes between the sugar-free group and sugar-containing group were analyzed with
linear regression: we considered P ≤ 0.05 significant.
b
Endpoints of 164 participants who withdrew from treatment were imputed, see Methods.
c
Completers were participants who consumed the drinks for the full 18 months. Mean changes may differ slightly from difference between means at 18 and 0 months because a few participants lacked
measurements at either timepoint.
d
We calculated z scores of body-mass index and height from the Dutch 2009 reference data.2
Completers only c
Full cohort, noncompleters imputed b
Primary endpoint
BMI z scored
Secondary endpoints
Sum of four skinfolds (mm)
Waist-to-height ratio (%)
Electrical-impedance fat mass (kg)
Electrical-impedance fat mass (% of body weight)
Other endpoints
Weight (kg)
Height (cm)
Height z scored
Waist circumference (cm)
0 months
Table S4.5 Linear regression analyses with correction for baseline differences.a
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
97
CHAPTER 4
Table S4.6 Adverse events.
a
Adverse Events
Treatment
Sugar-free group
Sugar-containing group
NonNonCompleters completers Total Completers completers Total
Headache
0
0
0
1
1
2
Allergy
0
2
2
0
3
3
Weight increase
0
2
2
2
4
6
Weight decrease
1
0
1
0
0
0
Behavioral problems
0
2
2
1
0
1
Abdominal discomfort
0
2
2
2
5
7
Total
1
8
9
6
13
19
a
We urged parents to report adverse events at the email address or cell phone number printed
on all beverage cans and gave all parents the telephone number of an independent physician
to report and discuss adverse events. None of the parents approached this physician. Adverse
events were reported by 21 noncompleters as a reason to stop drinking the beverages and by 7
children who completed the study.
98
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
a
Table S4.7 Baseline characteristics of completers versus noncompleters.
Completers Noncompleters
b
Characteristic
(n = 477)
(n = 164)
Girls
43%
57%
Age (years)
8.13±1.84
8.38±1.88
c
Dutch ancestry
87%
79%
Non-western ancestry
12%
20%
d
Parent education lower to intermediate
11%
20%
Parent education high-school
31%
29%
Parent education college or university
57 %
49%
Weight (kg)
29.76±8.30
31.43±10.27
Height (cm)
132.1±12.5
133.9±12.9
f
Height z score (SD units above Dutch mean)
−0.06±0.97
0.04±1.05
e
Body-mass index
16.73±2.29
17.15±3.23
f
Body-mass index z score (SD units above Dutch mean)
0.01±0.99
0.08±1.09
g
Low body-mass index
1%
1%
Healthy body-mass index
80%
81%
Overweight
17%
12%
Obese
2%
7%
h
Sum of four skinfolds (mm)
35.12±16.36
38.68±21.17
Waist-to-height ratio (%)
44.33±3.67
44.60±4.89
i
Electrical-impedance fat mass (kg)
5.53±3.32
6.41±4.75
Electrical-impedance fat mass (% of body weight)
18%
19%
a
Differences between completers and noncompleters at baseline were significant for Sex,
Ancestry, Weight status and Electrical-impedance fat mass, based on Chi-square and t-tests.
Plus-minus values are mean±SD.
b
94 children dropped-out in the sugar-free and 70 in the sugar group (P=0.03 for difference
between groups). For reasons to discontinue treatment, see Figure 4.1 of main paper and Table
S4.6 of the Supplementary Appendix.
c
N = 633; 8 households did not fill out this form. A child is designated Dutch if both parents
were born in the Netherlands, and Non-western if one or both parents were born in Suriname,
Dutch Antilles, Turkey, Morocco, Russia, Egypt or Vietnam.
d
N = 632; 9 households did not fill out this form. Lower to intermediate education is
Elementary school, Lower vocational secondary education, Technical secondary education,
Intermediate secondary education or Intermediate vocational education. We based
educational level on whichever of the parents had the highest education.
e
Body-mass index is the weight in kilograms divided by the square of the height in meters.
f
We calculated z scores of body-mass index and height from the Dutch 2009 reference data.2
g
We used international cut-offs for low and healthy body-mass index 4 and for overweight and
obesity.5
h
N = 640; 1 child refused the measurements.
I
N = 637; 4 children refused the measurements
99
CHAPTER 4
Table S4.8 Responses of participants to the question which type of beverage they
a
thought they had consumed.
Treatment
Response
‘Sweetened
with artificial
sweeteners’
‘Sweetened
with sugar’
‘I don't
know’ or
equivalent
Subtotals
Sugar-free group
81 (36%)
34 (15%)
107 (48%)
222
Sugar-containing group
67 (27%)
60 (24%)
125 (50%)
252
148
94
232
474
Sugar-free group
21 (27%)
12 (15%)
45 (58%)
78
Sugar-containing group
12 (21%)
11 (19%)
34 (60%)
57
33
23
79
135
Completers
Total
Noncompleters
Total
Total
609
a
Values are numbers of children, with percentage of subtotal in parentheses. At the final
visit at 18 months we asked available children (in Dutch): ‘Alle kinderen hebben dezelfde
smaken limonade gedronken: perzik, framboos, mango en citroen. De limonades die door
de kinderen werden gedronken waren allemaal zoet. Sommige kinderen hebben altijd
limonade gedronken dat zoet was gemaakt met suiker. Andere kinderen hebben altijd
limonade gedronken dat zoet was gemaakt met kunstmatige zoetstoffen. Welke limonade
heb jij gedronken: limonade met suiker of limonade met kunstmatige zoetstof?’ English
translation: ‘All children have consumed nonfizzy softdrinks with the same tastes: peach,
raspberry, mango and lemon. All the drinks that the children drank were sweet. Some
children always drank drinks that had been sweetened with sugar. Other children always
drank drinks that had been sweetened with artificial sweeteners. Which drink did you
consume: drinks sweetened with sugar or drinks with artificial sweetener?’
100
−1.05 (−1.73 to −0.37)
−0.97 (−1.52 to −0.42)
−0.36 (−0.71 to −0.02)
−0.07 (−0.14 to 0.01)
−0.69 (−1.22 to −0.17)
Weight (kg)
Height (cm)
Height z score
Waist circumference (cm)
0.01
0.08
0.04
0.001
0.003
0.001
0.02
0.003
0.001
P for
difference
−0.71 (−1.24 to −0.17)
−0.08 (−0.15 to −0.00)
−0.36 (−0.69 to −0.02)
−1.02 (−1.47 to −0.56)
−0.47 (−1.08 to 0.13)
−0.60 (−0.92 to −0.27)
−0.4 (−0.76 to −0.03)
−2.8 (−4.45 to −1.13)
−0.13 (−0.20 to −0.05)
Ancova-complete case
analysis with covariate
adjustment
Difference in changes
(95% CI)
0.01
0.05
0.04
< 0.001
0.12
< 0.001
0.03
0.001
0.001
−0.66 (−1.23 to −0.09
−0.04 (−0.10 to 0.02)
−0.37 (−0.72 to −0.02)
−1.01 (−1.54 to −0.48)
−1.07 (−1.99 to −0.15)
−0.57 (−1.02 to −0.12)
−0.4 (−1.0 to −0.0)
−2.2 (−4.0 to −0.4)
−0.13 (−0.21 to −0.05)
Independent sample t-test
after multiple imputation of
missing values
P for
Difference in changes
difference
(95% CI)
0.02
0.17
0.04
< 0.001
0.02
0.02
0.05
0.02
0.001
P for
difference
Outcomes of the independent sample t-test and the independent sample t-test after multiple imputation are copied from Table 4.2 of the main paper, and are given
here for comparison only. We included age at baseline, sex, ethnicity, and parental education as covariates in the Ancova-complete case analysis with covariate
adjustment.
a
−0.55 (−0.89 to −0.21)
Electrical-impedance fat mass
(% of body weight)
Other endpoints
−0.4 (−0.8 to −0.1)
Electrical-impedance fat mass (kg)
−2.5 (−4.2 to −0.8)
Waist-to-height ratio (%)
−0.13 (−0.20 to −0.06)
Sum of four skinfolds (mm)
Secondary endpoints
BMI z score
Primary endpoint
Difference in changes
(95% CI)
Independent sample t-test
Table S4.9 Comparison of outcomes according to the Independent sample t-test, Ancova-complete case analysis with covariate adjustment,3
and the Independent sample t-Test with values of noncompleters imputed.a
RESULTS OF THE DRINK TRIAL ON BODY WEIGHT
101
CHAPTER 4
References
1. de Ruyter JC, Olthof MR, Kuijper LD, Katan MB. Effect of sugar-sweetened
beverages on body weight in children: design and baseline characteristics of the
Double-blind, Randomized INtervention study in Kids. Contemp Clin Trials
2012;33:247-57.
2. Schonbeck Y, Talma H, van Dommelen P, et al. Increase in prevalence of
overweight in dutch children and adolescents: a comparison of nationwide growth
studies in 1980, 1997 and 2009. PLoS One 2011;6:e27608.
3. Groenwold RH, Donders AR, Roes KC, Harrell FE, Jr., Moons KG. Dealing with
missing outcome data in randomized trials and observational studies. Am J Epidemiol
2012;175:210-7.
4. Cole TJ, Flegal KM, Nicholls D, Jackson AA. Body mass index cut offs to define
thinness in children and adolescents: international survey. BMJ 2007;28:166-7.
5. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for
child overweight and obesity worldwide: international survey. BMJ 2000;320:1240-3.
102
CHAPTER 5
The effect of sugar-free versus sugar-sweetened
beverages on satiety, liking and wanting:
An 18 month randomized double-blind trial in children
Janne C de Ruyter
Margreet R Olthof
Lothar D J Kuijper
Djin G Liem
Martijn B Katan
Under review
CHAPTER 5
Abstract
Background
Substituting sugar-free for sugar-sweetened beverages reduces weight gain. A
possible explanation is that sugar-containing and sugar-free beverages cause the
same degree of satiety. However, this has not been tested in long-term trials.
Methods
We randomized 203 children aged 7-11 years to receive 250 mL per day of an
artificially sweetened sugar-free beverage or a similarly looking and tasting sugarsweetened beverage. We measured satiety on a 5-point scale by questionnaire at 0, 6,
12 and 18 months. We calculated the change in satiety from before intake to 1 minute
after intake and 15 minutes after intake. We then calculated the odds ratio that
satiety increased by 1 point in the sugar-group versus the sugar-free group. We also
investigated how much the children liked and wanted the beverages.
Results
146 children or 72% completed the study. We found no statistically significant
difference in satiety between the sugar-free and sugar-sweetened group; the adjusted
odds ratio for a 1 point increase in satiety in the sugar group versus the sugar-free
group was 0.77 at 1 minute (95% confidence interval, 0.46 to 1.29), and 1.44 at 15
minutes after intake (95% CI, 0.86 to 2.40). The sugar-group liked and wanted their
beverage more than the sugar-free group, adjusted odds ratio 1.63 (95% CI 1.05 to
2.54) and 1.65 (95% CI 1.07 to 2.55), respectively.
Conclusion
Sugar-sweetened and sugar-free beverages produced similar satiety. Therefore when
children are given sugar-free instead of sugar-containing drinks they might not make
up the missing calories from other sources. This may explain our previous observation
that children in the sugar-free group accumulated less body fat than those in the
sugar group.
106
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
Introduction
Recent trials have shown that sugar-free beverages lead to less weight gain than
sugar-sweetened drinks.1,2 A possible explanation is that sugars in solution are sensed
incompletely by receptors that determine satiation. As a result, sugar-free and sugarcontaining drinks should produce similar degrees of satiety,3 and intake of calories
from other foods is not affected.4 Previous studies indeed showed that masked
replacement of sugar-containing beverages with sugar-free beverages was only partly
compensated for by an increased energy intake.5,6 Small experiments lasting a few
days,7-9 or a few weeks10,11 showed that satiety following sugar-containing and sugarfree artificially sweetened drinks was similar. However, data from large, long-term
double blind trials are lacking, especially in children.
Liking and wanting are food rewards that also influence food intake. Liking
reflects the immediate experience or anticipation of pleasure from eating a food, i.e.
the hedonic value or ‘palatability’ of the food. Wanting is the intrinsic motivation to
engage in eating a food, now or in the (near) future.12 A four-week study in adults
found that sugar-free beverages were preferred less than sugar-containing
beverages,10 but another study failed to confirm this.11 Again, large long-term trials
are lacking
We therefore investigated satiety, liking and wanting in DRINK, a double-blind
randomized controlled trial in which children replaced their habitual daily sugarcontaining drink with either a sugar-free or sugar-sweetened beverage for 18
months.1,13
Methods
Ethics statement
Written informed consent was provided by a parent or guardian who had obtained
assent from the child. The Medical Ethical Committee of VU University Medical Centre
Amsterdam approved the study protocol.
Study population
The primary objective of the DRINK trial was to examine the effect of masked
replacement of sugar-sweetened beverages with noncaloric, artificially sweetened
beverages on weight gain. The design and results have been reported.1,13 Here we
report effects on satiety, liking and wanting. DRINK was an 18-month double-blind
107
CHAPTER 5
randomized controlled trial in 641 children aged 5-11 living near Amsterdam, the
Netherlands. Participants were individually randomized to receive either 250 ml per
day of a sugar-free, artificially sweetened beverage with 0 kcal (sugar-free group) or a
similar sugar-sweetened beverage that provided 104 kcal (sugar group). For the
present study, only participating children in the two highest school grades were
eligible because young children would not be able to understand the questionnaire.
All 203 of them proved willing and able to participate, and were enrolled (Fig. 5.1).
We recruited participants between August and November 2009.13 Treatment started
between November 14 and December 7, 2009. The trial lasted 19.5 months. We
interrupted treatment for 1.5 months during the summer holidays of 2010. Written
informed consent was provided by a parent or guardian who had obtained assent
from the child. The Medical Ethical Committee of VU University Medical Centre
Amsterdam approved the study protocol.
Study beverages
Dutch primary school children habitually bring a snack and a beverage to school for
the morning break. We replaced the beverage brought from home with our study
beverages. We provided children with 1 can per day of a noncaloric, artificially
sweetened, noncarbonated beverage or a sugar-sweetened noncarbonated beverage.
We developed custom drinks for this study to ensure that the sugar-free and sugarsweetened drinks tasted and looked essentially the same. The identical-looking 250ml cans provided either 0 or 26 g of sucrose (0 or 104 kcal per day). The sugar-free
beverages contained 34 mg sucralose and 12 mg acesulfame potassium as
sweeteners. Each week, children received a box at school labeled with their name and
containing 8 cans, 1 for each day of the week plus 1 extra to be used as a spare in case
a can was misplaced. We offered beverages in four flavors: raspberry, peach, lemon
and mango. Flavors were rotated every two weeks.
Satiety, liking and wanting
We measured satiety, liking and wanting with a questionnaire (Fig. 5.2;
Supplementary Appendix). The questions about satiety14 and liking15 have previously
been validated in children in this age group, and the question about wanting has also
been used earlier for research in children.16 Children completed the questionnaire
during their 10:00 am morning break, during which they consumed their study drink
together with the snack that they habitually brought from home. Measurements were
108
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
641 Randomized for DRINK main study
438 in grades 2,3 and 4 excluded
203 in grade 5 and 6 eligible
203 in grade 5 and 6 included
100 sugar-free group
103 sugar group
11 discontinued study before 6-mo
6 discontinued study before 6-mo
measurement
measurement
9 No longer liked the beverage
3 No longer liked the lemonade
1 Had adverse events
2 Had adverse events
1 Declined to participate further
1 Declined to participate further
10 discontinued study between 6-mo and
7 discontinued study between 6-mo and 12-
12 mo measurement
mo measurement
6 No longer liked the beverage
7 No longer liked the lemonade
4 Declined to participate further
13 discontinued study between 12-mo and
10 discontinued study between 12-mo and
18-mo measurement
18-mo measurement
12 No longer liked the lemonade
9 No longer liked the lemonade
1 Declined to participate further
1 Moved
66 completed the study
80 completed the study
Figure 5.1 Randomization and follow-up of the study participants.
made at the start of the study (0 months), and again at 6, 12 and 18 months into the
study. The teachers instructed the children how to fill out the questionnaire, and one
member of our research team was also present to assist if necessary. We measured
109
CHAPTER 5
satiety, liking and wanting on 5-point scales; 1 indicated low levels, and 5 high levels
(Supplementary Appendix). We measured satiety three times on each test day: just
before intake of the study beverage and the food brought from home, 1 minute after
intake, and 15 minutes after intake.3 We measured wanting 1 minute before and
liking 1 minute after intake. We also asked whether children consumed the entire
content of the can and what type of food and how much they ate together with the
study beverage. We calculated the calories in the food.17,18
9:59– 10:00 am
10:00– 10:15 am
10:15– 10:16 am
10:16– 10:31 am
10:31– 10:32 am
Score wanting
Eat snack and drink
beverage
Score liking
Leisure time
Score satiety 15 minutes
after intake
Score satiety before intake
Record type and amount
of snack
Score satiety 1 minute
after intake
Record whether entire
contents of can were
consumed
Figure 5.2 Time schedule for filling out the questionnaire. The timeline is indicative
because each school had its own timing of the morning break.
Statistical Analyses
We present outcome variables as medians and interquartile ranges (Table 5.2). We
used generalized linear mixed model analysis in STATA SE 12 (StataCorp LP, Texas,
USA). We calculated regression coefficients for the likelihood that satiety increased
from before intake to 1 minute and 15 minutes after intake by 1 point on the 5-point
scale in the sugar-group versus the sugar-free group (Table 5.3a). We adjusted
calculations of satiety after for satiety prior to beverage intake (model 1). We also
adjusted for caloric intake from food with the beverage, and for gender and BMI z
score to correct for imbalances in randomization (model 2). We used similar analyses
to calculate differences in liking and wanting of the beverages between treatment
groups (Table 5.3b). Calculations of liking and wanting were adjusted for gender and
BMI z score at the start of the study. We also calculated regression coefficients for
effect modification by month of measurement. For the analyses for satiety, liking and
wanting, the sugar-free group was the reference group. We used the conventional
limits for significance of P = 0.10 for effect modification, and P = 0.05 for all other
analyses.
110
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
Results
Participants
The 203 participants were aged 10.2±0.8 years (mean±SD) at the start of the study.
Baseline characteristics were similar for the two treatment groups, except that the
sugar group had more girls, a lower BMI, and parents had achieved higher education
levels (Table 5.1). A total of 146 (72%) children completed the study, 66 in the sugarfree group and 80 in the sugar group (Fig. 5.1). Most children (81%) who stopped
drinking the beverages did so because they no longer liked the beverages.
Study beverage and food intake
The percentage of children who consumed the entire content of the can on the
morning when questionnaires were applied ranged from 87% to 98% over the 4
months of measurement. These percentages were similar for the two groups (Table
5.2). The caloric intake from the snacks brought from home was 137±90 kcal
(mean±SD) in the sugar-free group and 140±84 kcal in the sugar group (Table 5.2).
Most children (77%) brought sweet snacks such as crackers, cookies, cakes, or bars,
20% brought fruits, 2% brought savory snacks such as cheese and chips, and 1%
brought bread.
Satiety
In both groups, satiety was lowest before and highest 1 minute after beverage intake,
and then decreased during the subsequent 14 minutes (Table 5.2; Fig. 5.3). Sugar
content of the beverages did not statistically significantly affect satiety (Table 5.3a;
Fig. 5.4). At 1 minute after intake, the odds ratio was 0.84 which implies that the
likelihood that satiety increased by 1 point was 0.84 times as high in the sugar group
as in the sugar-free group (Table 5.3a). At 15 minutes, the effect was in the opposite
direction, with an odds ratio of 1.49, but again this was not statistically significant.
Adjustments for caloric intake from the snack, gender and BMI z score at the start of
the study had only minor effects on outcomes (Table 5.3a). Effect modification by
timepoint, i.e. month of measurement, was not statistically significant 1 minute after
intake (P = 0.53), but marginally significant 15 minutes after intake (P=0.09).
111
CHAPTER 5
Table 5.1 Baseline characteristics of participants.
Characteristic
a
Sugar-free
group
(N = 100)
42%
10.2±0.73
89%
10%
20%
23%
56 %
36.86±6.14
144.4±7.28
17.6±2.2
0.02±0.99
6%
78%
16%
0%
41.5±16.7
43.6±3.9
7.5±3.2
20%
Sugar group
(N = 103)
Girls (%)
52%
Age (years)
10.2±0.80
b
Dutch ancestry
92%
Non-western ancestry
7%
c
Parent education lower to intermediate
8%
Parent education high-school
25%
Parent education college or university
66%
Weight (kg)
36.36±6.51
Height (cm)
145.5±7.49
d
Body-mass index
17.1±2.2
e
Body-mass index z score (SD units above Dutch mean)
-0.09±0.90
f
Low body-mass index
3%
Healthy body-mass index
85%
Overweight
12%
Obese
0%
Sum of four skinfolds (mm)
37.8±17.4
Waist-to-height ratio (%)
42.3±3.6
g
Electrical-impedance fat mass (kg)
6.9±3.3
Electrical-impedance fat mass (% of body weight)
18%
a
Values are means ±SD or percentages, as indicated.
b
N = 201; 2 households did not fill out this form. A child is designated Dutch if both
parents were born in the Netherlands, and Non-western if one or both parents were born
in Suriname, Dutch Antilles, Turkey, Morocco, Russia, Egypt or Vietnam.
c
N = 201; 2 households did not fill out this form. Lower to intermediate education is
Elementary school, Lower vocational secondary education, Technical secondary education,
Intermediate secondary education or Intermediate vocational education. We based
educational level on whichever of the parents had the highest education.
d
Body-mass index is the weight in kilograms divided by the square of the height in meters.
e
We calculated z score of body-mass index and height from the Dutch 2009 reference
29
data.
f
30
We used international cut-offs for low and healthy body-mass index and for overweight
31
and obesity.
g
N = 202; 1 child refused the measurements
112
5(4-5)
4(4-5)
4(3-4)
4(3-4)
2(1-3)
3(3-4)
2(2-3)
N=86
6 Mo
Sugar-free
Group
4(3-4)
4(3-5)
2(1-2)
3(2-4)
2(2-3.5)
3.5(3-4)
3(2-4)
2(1-2)
3(2-4)
2(1-3)
N=75
3(3-4)
3(2.25-4)
2(1-2.75)
3.5(2-4)
3(2-4)
N=88
12 Mo
Month of measurement
6 Mo
12 Mo
N=95
Sugar
Group
Sugar-free
Group
Sugar
Group
3(2-4)
3(2-4)
2(1-3)
3(2-4)
3(2-4)
N=63
18 Mo
Sugar-free
Group
3(2-4)
3(2-4)
2(1-2)
3(2-4)
3(2-4)
N=78
18 Mo
Sugar
Group
Percentage of children who drank entire can
95%
98%
98%
97%
87%
89%
93%
88%
Snack intake (kcal) consumed with the study
129.2
131.9
149.6
144.4
127.4
139.7
142.9
147.2
beverage, mean (±SD)
(±81.9)
(±77.3)
(±102.3)
(±87.7)
(±76.2)
(±87.0)
(±97.3)
(±83.4)
a
Values are median (interquartile range) unless otherwise noted. Numbers of participants do not match those in Figure 5.1 because absentees.
4(3-5)
4(4-5)
Liking of the beverage
Wanting of the beverage
N=99
N=95
2(1-2)
3(2-4)
2(2-3.25)
0 Mo
0 Mo
1(1-2)
3(2-4)
2(1-3)
Sugar
Group
Sugar-free
Group
Satiety before intake
Satiety 1 minute after beverage intake
Satiety 15 minutes after beverage intake
Variable
Table 5.2 Medians with interquartile ranges of satiety, liking, wanting, and beverage- and snack intake, measured on a 5-point scale
in children.a
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
113
CHAPTER 5
Panel B
After 15
minutes
Figure 5.3 Satiety after intake minus satiety before intake of the beverages at 18 months
in 146 children. Bars indicate the shift in satiety from before to after intake of the
beverages. Panel A shows satiety 1 minute after intake; Panel B, 15 minutes after intake.
An increase of 1 means any increase of 1 point on the 5-point scale.
Liking and wanting
The children randomized to receive sugar-sweetened beverages liked these more
than those who received sugar-free beverages for 18 months; the odds ratio was 1.58
(Table 5.3b). However, when the scores on the 5-point scale were averaged
arhythmetically the differences between the group means were small, and were much
less than the overall fall in liking over the course of the trial (Fig. 5.4). Average liking
fell by 1.18 points in the sugar group and by 1.04 points in the sugar-free group, while
the difference between groups was 0.4±0.2 points (mean±SE) at baseline and 0.2±0.2
points at 18 months (Fig. 5.4). Effect modification by month of measurement was not
statistically significant (P= 0.27). This suggests that duration of consuming the drinks
did not influence the differences in liking between groups. The children also wanted
sugar-sweetened beverages more than sugar-free beverages; the odds ratio was 1.59
(Table 5.3b; Fig. 5.4). Again, effect modification by month of measurement was not
statistically significant for differences in wanting between groups (P= 0.97).
114
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
Figure 5.4 Arhythmetic mean scores for satiety, liking, and wanting over the course of
the trial. Dashed lines, sugar group; solid lines, sugar-free group. T bars indicate one
standard deviation
Discussion
We found that satiety following sugar-free and sugar-sweetened beverages was the
same in a large population of children who consumed such beverages daily for 18
months. We previously reported that children who consumed sugar-free beverages
accumulated less body fat than children who consumed sugar-sweetened beverages.1
The present study suggests that this may be explained by similar levels of satiety.
When we substituted sugar-free beverages for the sugar-containing drinks that
children drank habitually, they apparently did not feel a need to compensate for the
missing calories by eating more of other foods and drinks.
We tested satiety under real-life conditions, i.e. during the morning break at
school when children consumed their beverage together with their usual snack. We
assume that absolute satiety after 1 and 15 minutes was determined by snack intake,
beverage volume, and caloric content of the drink. Mean caloric intake from snacks
and volume of drinks were similar between groups but the sucrose content of the
beverages differed. We therefore conclude that the sugar content of the drinks did
115
CHAPTER 5
not have a major effect on satiety. This finding is consistent with short term
experiments in adults7-11 that also found similar levels of satiety following sugarcontaining and sugar-free beverages. There was some indication to suggest that
sugar-sweetened beverages became more satiating as the trial proceeded because
effect modification by month of measurement for 15 minutes after intake was
marginally statistically significant. However, this may have been a chance finding.
We found that the children liked and wanted the sugar-sweetened beverage
slightly more than the sugar-free beverage. These differences persisted throughout
the study even though overall liking and wanting of both types of beverage fell
drastically with time. In contrast, short-term studies with pudding,19 or beverages11
found similar ratings of pleasantness for aspartame-sweetened and sucrosesweetened products. However, beverages containing a blend of aspartame,
acesulfame potassium and saccharin were rated lower in pleasantness than beverages
containing sucrose.10 Another study reported that cream cheese sweetened with
aspartame had a more pleasant taste than cream cheese containing stevia or
sucrose.20 Thus the pleasantness of artificially sweetened products may be highly
dependent on the type and mix of sweeteners used and on other aspects of product
formulation.
Liking and wanting decreased markedly in both groups over the course of the
trial. Similar declines have been reported in studies with solid foods over periods of 15
days to 6 months.21-23 The decrease in liking and wanting agrees with our observation
that most children who discontinued intake gave dislike of the beverage as their
reason. A year and a half is indeed a long time for a child to drink the same drink day
after day despite the variation in flavors that we offered. Sensory-specific satiety may
have decreased liking and wanting over time.24
Our study had several strengths. One is its long duration. Previous studies had a
maximum duration of 4 weeks.10 Long term studies may be more informative because
short term satiety signals may have little to do with the long term mechanisms that
determine weight gain. Our study was large; we included 203 subjects, as opposed to
11 to 42 in previous studies.7-11 Also, we used a double-blind design that eliminated
the effects of psychological cues and socially desirable behavior. Previous studies
were either incompletely blinded7,9 or not blinded.8 Finally, we performed our study
in children while previous studies investigated adults only. Regulation of food intake
in adults may differ from that in children.25,26
Our study also had limitations. Of the 203 participants, 28% of children did not
complete the study. However, we found similar results in completers only (data not
116
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
shown). The lack of a statistically significant effect does not exclude that we may have
failed to pick up small differences in satiety between beverage groups. However, since
the effects were in opposite directions at 1 and 15 minutes after intake, we consider it
likely that sugar-free and sugar-sweetened beverages did not lead to systematically
different levels of satiety. The dietary status of our children was less standardized
than in previous studies, which included an overnight fast plus a standardized
breakfast.8,9 Finally, we measured satiety by questionnaire and did not quantitate
actual food intake following the beverage intake.7-9
The participants in our study were healthy Dutch children, most of whom were
white and of normal weight. Future studies should be done to find out whether the
long term effects of satiety would be similar in other ethnic groups, or obese children.
Future studies may also examine the effect on satiety with intervals longer than 15
minutes.27 One important question is whether our findings are unique for sugars in
liquid form. Short term studies with semi-solid foods such as pudding and jelly
suggested that noncaloric sweeteners produced the same degree of satiety as
sucrose.19,28 Studies with solid foods are lacking.
We found that sugar-sweetened and sugar-free beverages produced similar
satiety. Therefore when children are given sugar-free instead of sugar-containing
drinks they might not make up the missing calories from other sources. This may
explain our previous observation that the children in the sugar-free group
accumulated less body fat than those in the sugar group.1
Acknowledgements
We thank the children, parents, school administrators and teachers for their
cooperation; Emilie de Zoete and Hetty Geerars for assistance in the execution of the
study, and Michiel de Boer for statistical analyses.
117
CHAPTER 5
References
1. de Ruyter JC, Olthof MR, Seidell JC, Katan MB. A trial of sugar-free or sugarsweetened beverages and body weight in children. N Engl J Med 2012;367:1397-406.
2. Ebbeling CB, Feldman HA, Chomitz VR, et al. A randomized trial of sugarsweetened beverages and adolescent body weight. N Engl J Med 2012;367:1407-16.
3. Blundell J, de Graaf C, Hulshof T, et al. Appetite control: methodological aspects
of the evaluation of foods. Obes Rev 2010;11:251-70.
4. Cassady BA, Considine RV, Mattes RD. Beverage consumption, appetite, and
energy intake: what did you expect? Am J Clin Nutr 2012;95:587-93.
5. Tordoff MG, Alleva AM. Effect of drinking soda sweetened with aspartame or
high-fructose corn syrup on food intake and body weight. Am J Clin Nutr 1990;51:963-9.
6. Porikos KP, Booth G, Van Itallie TB. Effect of covert nutritive dilution on the
spontaneous food intake of obese individuals: a pilot study. Am J Clin Nutr
1977;30:1638-44.
7. Rolls BJ, Kim S, Fedoroff IC. Effects of drinks sweetened with sucrose or
aspartame on hunger, thirst and food intake in men. Physiol Behav 1990;48:19-26.
8. Holt SH, Sandona N, Brand-Miller JC. The effects of sugar-free vs sugar-rich
beverages on feelings of fullness and subsequent food intake. Int J Food Sci Nutr
2000;51:59-71.
9. Canty DJ, Chan MM. Effects of consumption of caloric vs noncaloric sweet drinks
on indices of hunger and food consumption in normal adults. Am J Clin Nutr
1991;53:1159-64.
10. Van Wymelbeke V, Beridot-Therond ME, de La Gueronniere V, Fantino M.
Influence of repeated consumption of beverages containing sucrose or intense
sweeteners on food intake. Eur J Clin Nutr 2004;58:154-61.
11. Beridot-Therond ME, Arts I, Fantino M, De La Gueronniere V. Short-term effects
of the flavour of drinks on ingestive behaviours in man. Appetite 1998;31:67-81.
12. Mela DJ. Eating for pleasure or just wanting to eat? Reconsidering sensory
hedonic responses as a driver of obesity. Appetite 2006;47:10-7.
13. de Ruyter JC, Olthof MR, Kuijper LD, Katan MB. Effect of sugar-sweetened
beverages on body weight in children: design and baseline characteristics of the
Double-blind, Randomized INtervention study in Kids. Contemp Clin Trials
2012;33:247-57.
14. Faith MS, Kermanshah M, Kissileff HR. Development and preliminary validation
of a silhouette satiety scale for children. Physiol Behav 2002;76:173-8.
118
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
15. Léon F, Couronne T, Marcuz MC, Köster EP. Measuring food liking in children: a
comparison of non verbal methods. Food Quality and Preference 1999;10:93-100.
16. Liem DG, Zandstra LH. Children's liking and wanting of snack products: Influence
of shape and flavour. Int J Behav Nutr Phys Act 2009;6:38.
17. RIVM. NEVO-online versie 2011/3.0. http://nevo-online.rivm.nl. Accessed
November 15, 2012.
18. Donders-Engelen M, Van der Heijden L, Hulshof K. Maten, Gewichten en
Codenummers 2003 (English translation: Units, Weights and Codes 2003)
Wageningen: Wageningen University, TNO; 2003.
19. Rolls BJ, Laster LJ, Summerfelt A. Hunger and food intake following consumption
of low-calorie foods. Appetite 1989;13:115-27.
20. Anton SD, Martin CK, Han H, et al. Effects of stevia, aspartame, and sucrose on
food intake, satiety, and postprandial glucose and insulin levels. Appetite 2010;55:37-43.
21. Zandstra EH, de Graaf C, van Trijp HC. Effects of variety and repeated in-home
consumption on product acceptance. Appetite 2000;35:113-9.
22. Rolls ET, de Waal AW. Long-term sensory-specific satiety: evidence from an
Ethiopian refugee camp. Physiol Behav 1985;34:1017-20.
23. Hetherington MM, Bell AC, Rolls BJ. Effects of repeat consumption on
pleasantness, preference and intake. British Food Journal 2000;102:507-21.
24. Rolls BJ, Rolls ET, Rowe EA, Sweeney K. Sensory specific satiety in man. Physiol
Behav 1981;27:137-42.
25. Birch LL, Deysher M. Conditioned and unconditioned caloric compensation:
Evidence for self-regulation of food intake in young children. Learning and Motivation
1985;16:341-55.
26. Bellisle F, Drewnowski A, Anderson GH, Westerterp-Plantenga M, Martin CK.
Sweetness, satiation, and satiety. J Nutr 2012;142:1149S-54S.
27. Erlanson-Albertsson C. Appetite regulation and energy balance. Acta Paediatr
Suppl 2005;94:40-1.
28. Rolls BJ, Hetherington M, Laster LJ. Comparison of the effects of aspartame and
sucrose on appetite and food intake. Appetite 1988;11 Suppl 1:62-7.
29. Schonbeck Y, Talma H, van Dommelen P, et al. Increase in prevalence of
overweight in dutch children and adolescents: a comparison of nationwide growth
studies in 1980, 1997 and 2009. PLoS One 2011;6:e27608.
30. Cole TJ, Flegal KM, Nicholls D, Jackson AA. Body mass index cut offs to define
thinness in children and adolescents: international survey. BMJ 2007;28:166-7.
119
CHAPTER 5
31. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for
child overweight and obesity worldwide: international survey. BMJ 2000;320:1240-3.
120
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
Supplementary Appendix
Sensory questionnaire in Dutch (English translation in italics)
121
CHAPTER 5
VOORNAAM (Name):
ACHTERNAAM (Last name):
GROEP (Grade):
SCHOOL (School):
VRAGENLIJST
(Questionnaire)
122
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
Hoi!
Voor je ligt de DRINK-studie vragenlijst. De vragen
gaan over wat je van Blikkie vindt. Bij het invullen gaat
het erom dat je laat weten hoe JIJ je voelt. Er is hierbij
geen goed of fout. Als er staat dat je Blikkie op mag
drinken, dan eet en drink je rustig zoals je gewend bent,
voordat je de volgende vragen invult.
Als je bij de RODE BLADZIJDE bent, STOP je met
invullen en na de pauze vul je de laatste bladzijde in. Na
het invullen kan je de vragenlijst bij je juf of meester
inleveren.
Alvast bedankt!
Hi!
In front of you is the DRINK study questionnaire. The questions
are about what you think of Blikkie. When you fill in the
questionnaire we would like to know how you are feeling. There
are no incorrect or correct answers. When you come across
‘Time to eat & drink’, you first eat and drink as you normally
would. You then fill in the questions.
When you see a red page, stop. You now go outside, and after the
break, you fill in the last question.
When you are finished, please give the questionnaire to the
teacher.
Thanks
123
CHAPTER 5
Hoeveel zin heb je in het drankje?
(Do you feel like drinking the beverage?)
Voor dat je Blikkie proeft willen wij graag weten hoeveel zin je NU in
BLIKKIE hebt. Je kruist een hokje aan achter het blikje waarvan jij denkt
dat bij jou past
(Before you taste Blikkie, we would like to know if you feel like drinking
Blikkie. You tick the box that matches what you feel)
Ik heb nu HELEMAAL
GEEN ZIN in Blikkie
(I really do not want to
drink Blikkie)
Ik heb nu GEEN ZIN in
Blikkie (I do not want
to drink Blikkie)
Ik heb nu EEN BEETJE
ZIN in Blikkie (I want to
drink Blikkie a little bit)
Ik heb nu BEST WEL
ZIN in Blikkie
(I want to drink Blikkie)
Ik heb nu HEEL VEEL
ZIN in Blikkie (I really
want to drink Blikkie)
124
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
Hoe vol zit je nu, VLAK VOORDAT je gaat eten en
drinken?
(How full do you feel, right before eating and drinking?)
Wij willen weten hoe vol jij je voelt. Hieronder zie je 5 plaatjes van
kinderen. Alle kinderen hebben een beetje eten in hun buik. Achter elk
plaatje staat hoe het kind zich voelt. Zet een kruisje achter het kind dat zich
net zo voelt als jij je NU voelt
(We would like to know how full you feel. You see five pictures with
children. All the children have some food in their stomach. Next to the
picture we wrote the feeling of that child. You tick the box that matches what
you feel)
Ik voel me HELEMAAL NIET VOL. Ik
zou nog heel veel kunnen eten (I am not
full at all. I could eat a lot)
Ik voel me EEN KLEIN BEETJE VOL.
Ik zou nog wel wat kunnen eten (I am a
little bit full. I could eat)
Ik voel me EEN BEETJE VOL. Ik zou
nog een beetje kunnen eten (I am a bit
full. I could eat a little bit)
Ik voel me BEST WEL VOL. Ik zou nog
een klein beetje kunnen eten (I am quite
full. I could eat only a very little bit)
Ik voel me HEEL VOL. Ik zou niets
meer kunnen eten (I am completely full.
I could not eat anymore)
125
CHAPTER 5
EET & DRINKPAUZE
(Time to eat & drink)
Nu mag je Blikkie opdrinken en gaan eten. We willen graag een paar dingen
weten over het eten en drinken. De eerste vraag mag je met ja of nee
beantwoorden
(Now its time to drink Blikkie and eat your snack. We would like to know a
few things about what you eat and drink. You can answer the first question
with yes or no).
Eet je bij je BLIKKIE? (Are you eating while drinking your Blikkie?)
Ja (yes)
Nee (no)
Als je antwoord Nee was, ga je nu naar de volgende bladzijde. Als je
antwoord Ja was, willen we nu precies weten wat je bij Blikkie eet. Je
mag je antwoord schrijven op de stippellijn
(If you answered with ‘no’, please go to the next page. If you answered with
‘yes’, we would like to know exactly what you are eating. Your can write
your answer on the dotted line).
Bijvoorbeeld (For example):
Wat eet je nu bij het drankje? Koekje
(What are you eating with the beverage? Cookie)
Hoeveel eet je ervan? 2 stuks
(How much are you eating? 2 pieces)
Welke naam staat er op de verpakking? Sultana naturel
(Which name is printed on the packaging? Sultana natural)
Wat eet je nu bij het Drankje?....................................
(What are you eating with the beverage?)
Hoeveel eet je ervan?....................................................
(How much are you eating?)
Welke naam staat er op de verpakking?....................
(Which name is printed on the packaging?)
126
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
Hoe lekker vind je Blikkie?
(How much do you like Blikkie?)
Nu je Blikkie opgedronken hebt, willen wij graag weten hoe lekker je
Blikkie vindt. Hieronder zie je 5 gezichtjes. Zet een kruisje achter het
gezichtje dat bij jou past. (Now that you finished Blikkie, we would like to
know how much you like Blikkie. You see five pictures with faces. You tick
the box that matches what you feel)
Ik vind Blikkie heel erg
lekker
(I find Blikkie delicious)
Ik vind Blikkie
lekker
(I like Blikkie)
Ik vind Blikkie niet vies
maar ook niet lekker
(I neither like nor
dislike Blikkie)
Ik vind Blikkie een
beetje vies (I do not like
Blikkie)
Ik vind Blikkie heel erg
vies (I find Blikkie
disgusting)
127
CHAPTER 5
Beantwoord nu de volgende vraag met ja of nee.
(Answer the following question with ‘yes’ or ‘no’)
Heb je BLIKKIE helemaal leeg gedronken?
(Did you finish the whole can?)
Ja (yes)
Nee (no)
128
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
Hoe vol zit je nu, DIRECT NADAT je gegeten
en gedronken hebt? (How full do you feel, directly after eating
and drinking?)
NU JE BLIKKIE OPGEDRONKEN HEBT, willen wij weten hoe vol jij
je voelt. Zet een kruisje achter het kind dat zich net zo voelt als jij je NU
voelt. (Now that you finished Blikkie, we would like to know how full you
feel. You tick the box that matches what you feel)
Ik voel me HELEMAAL NIET VOL.
Ik zou nog heel veel kunnen eten. (I am
not full at all. I could eat a lot)
Ik voel me EEN KLEIN BEETJE
VOL. Ik zou nog wel wat kunnen eten.
(I am a little bit full. I could eat)
Ik voel me EEN BEETJE VOL. Ik
zou nog een beetje kunnen eten. (I am
a bit full. I could eat a little bit)
Ik voel me BEST WEL VOL. Ik zou
nog een klein beetje kunnen eten. (I am
quite full. I could eat only a very little
bit)
Ik voel me HEEL VOL. Ik zou niets
meer kunnen eten. (I am completely
full. I could not eat)
129
CHAPTER 5
NU HEB JE PAUZE
(Break time)
130
RESULTS OF THE DRINK TRIAL ON SATIETY, LIKING AND WANTING
DIRECT NA DE PAUZE
(Right after the break)
Nu je pauze hebt gehad, willen we graag weten hoe vol jij je NU voelt. Zet
een kruisje achter het kind dat zich net zo voelt als jij je NU voelt. (Now that
the break is over, we would like to know how full you feel. You tick the box
that matches what you feel)
/;
Ik voel me HELEMAAL NIET VOL.
Ik zou nog heel veel kunnen eten. (I am
not full at all. I could eat a lot)
Ik voel me EEN KLEIN BEETJE
VOL Ik zou nog wel wat kunnen eten.
(I am a little bit full. I could eat)
Ik voel me EEN BEETJE VOL. Ik
zou nog een beetje kunnen eten.
(I am a bit full. I could eat a little bit)
Ik voel me BEST WEL VOL Ik zou
nog een klein beetje kunnen eten.
(I am quite full. I could eat only a very
little bit)
Ik voel me HEEL VOL. Ik zou niets
meer kunnen eten. (I am completely
full. I could not eat anymore)
131
CHAPTER 5
Dank je wel voor het invullen!
Je kunt deze vragenlijst nu inleveren bij je juf
of meester
(Thank you for filling in the questionnaire. Please give the questionnaire to
the teacher)
132
CHAPTER 6
General discussion
CHAPTER 6
The studies in this thesis focused on two major issues. We examined the effects of
masked replacement of sugar-sweetened beverages by sugar-free beverages on body
weight and fat accumulation. We also examined the effects of such drinks on sensory
aspects: satiety, liking and wanting. DRINK showed that children in the sugar-free
group increased less in body weight, and accumulated less body fat than those in the
sugar group, and that satiety of both types of beverages was similar. Children liked
and wanted the sugary beverage slightly more compared to the sugar-free beverage.
Methodological strengths and limitations
Strengths
Double blind design
A strength of the study was the double blind design of DRINK. This design eliminated
the effects of behavioral, cognitive, and psychological factors. We managed to
manufacturing sugar-sweetened beverages and sugar-free beverages that tasted and
looked the same. Earlier trials were not properly blinded.1-5 By using a double blind
design the results provided the clearest evidence as yet of how drinking sugarsweetened beverages can affect children’s weight gain with unconscious physiological
mechanisms. Although blinding was imperfect, it was more successful than in most
randomized, double-blind trials.6 A few more children in the sugar-free group than in
the sugar-group were aware of the content of the cans. However we do not think that
this has substantially influenced the outcomes of the study.
Sample size and study duration
DRINK included a large number of children. According to our power calculation, 212
subjects per treatment should be enrolled to detect a statistically significant
difference if there proved to be one.7 641 children were included, and 477 children
consumed the beverages for the full 18 months (225 in sugar-free group, and 252 in
sugar-group). This means that enough children finished the trial to answer the
research question with sufficient statistical power. Previous trials that included 23 to
318 subjects may therefore have been too small to detect an effect.2,3,5 Other studies
were larger, but in those studies children were not individually randomized.1,4 The
sensory questionnaire study included 203 subjects, as opposed to 11 to 42 in previous
studies.8-13 However, we did not perform a power calculation for the sensory
questionnaire study and included all eligible children who signed up.
The 18-month (78 weeks) study duration ensured that the observed effect was
not transient. Other trials on the effect of sugary beverages on body weight varied
136
GENERAL DISCUSSION
from 3 to 52 weeks.1-5 Earlier trials that examined the effect of sugary drinks on
satiety lasted a few days,9-11,13 or a few weeks,8,12 with a maximum of 4 weeks.
Stratification and randomization
Participants were stratified by school, gender, age and baseline BMI to ensure that
these potential determinants of weight gain were evenly distributed between the
treatment and control group. The children were then individually randomized to
balance other behaviors that may affect body weight. For example, it was likely that
both groups encompassed similar numbers of children who watched a lot of
television, ate a lot of candy, or were more or less active at baseline. The stratified
randomization produced well-balanced study groups at baseline. Therefore any
difference between both groups in the study outcomes could be attributed to the
beverages and not to other life style factors.
Limitations
Dropout
In the main trial 26% of the participants did not complete the study. It is unlikely that
selective dropout occurred because as long as dropouts were participating in the
study, their changes in weight and body fat paralleled those in children who
ultimately completed the study. Therefore the effect of the beverage consumed in the
study was not different in these children compared to those who completed the
study. Data on weight and height at the end of the study were available for most of
the dropouts. When the data of those measurements were pooled with the
measurements of the children who completed the study, the effect of the study
beverage became smaller and non-significant. This finding was expected; the children
who did not complete the study probably went back to drinking sugary beverages,
which attenuated the effect of the sugar-free beverages that they consumed before
discontinuation.
Food consumption
We could have attempted to measure what the children ate and drank in response to
the treatments. However, it is known that it is difficult to measure food intake in free
living subjects to the degree of precision required for a trial. The ideal way to measure
energy intake is the double labeled water method.14 However, to perform such
measurements was not feasible. As body fatness was the focus, BMI z score became
the primary endpoint. DRINK suggests that removal of sugary drinks was only partly
compensated for by eating or drinking extra as to compensate for the removal of
137
CHAPTER 6
sugar. However, data on food intake that show that children did indeed partly
compensate for the removal of calories are lacking.
Generalizability
The participants were healthy Dutch children, most of whom were white and of
normal weight. Thus, it is unknown whether the results would be similar in other
ethnic groups, or adults. However, it is most likely that the same biologic mechanisms
are operative.
Reflections & future research
Reflections
Is a change of 0.13 in BMI z score a meaningful outcome?
Previous research showed that children of parents that attained high levels of
education had a BMI z score that was 0.27 SD units lower than children of parents
with poor education.15 This means that the difference of 0.13 BMI z score between
the study groups was approximately half of the difference between children raised in
families with high or low educated parents. I therefore consider the effect that can be
achieved by the removal of sugar-sweetened drinks substantial.
What effect on body weight did we expect?
Predicting weight gain and loss in children is precarious. Attempts have been made;
however these numbers should be interpreted with caution. Naïve calculations
suggest that an increase of 3500 kcal results in 454 gram of weight increase.16 In
DRINK, children consuming sugar-free drinks had an estimated diminished intake
compared to their controls of 26 g sugar daily during 541 days. This corresponds with
a total of 56 264 kcal. This would then result into a weight difference of approximately
7 kg. However, this assumes that human bodies are static units in which calories
accumulate; this is not the case. Human bodies are dynamic systems in which basal
metabolic rate and energy costs of body activity go up when body weight increases.1719
For the power calculation of DRINK we used the model of Wang et al.17 This model
predicted a weight difference of 2.3 kg in the case of full compliance, and absence of
compensation.7 After we had already started DRINK, Hall et al.19 proposed a model to
predict weight changes in adults which forecast a weight difference after 541 days
with 85% compliance of about 2.1 kg. In DRINK we found a difference in body weight
of 1.01 kg or 0.83 kg after correction for height difference between the sugar and
sugar-free group. This was substantially lower than the 2.1 kg predicted by Hall et al.19
138
GENERAL DISCUSSION
This suggested that 52% compensation must have occurred; meaning that for 52% of
the removed calories children ate or drank extra. A review estimated that in adults
the removal of sugar is compensated for by 16-32%.20 The true value probably lies
somewhere in between 16% and 52%. However, these are the numbers for adults not
for growing children. It is more difficult to construct models for children because they
cannot be properly validated due to lack of experimental data on weight loss or gain
in children. However, the actual numbers for children are less relevant for policy
decisions. What matters is that removal of a single sugary beverage daily during an 18
month period reduced body weight gain with 1.01 kg or 0.83 kg after correction for
height difference.
What is the effect of masked replacement of sugar-sweetened by sugar-free
beverages on body weight in overweight and obese children?
DRINK was not designed to examine whether switching sugar-free drinks would
prevent body weight gain in obese children. However, 19% (46 children in sugar-free
group, and 45 children in sugar-group) of the DRINK participants who completed the
study were overweight or obese at baseline.21 We found that in this subgroup of
overweight and obese children, the sugar-free group increased 2.43 kg less in body
weight compared to the sugar group whereas this was 0.97 kg for those with a
healthy or lean BMI at baseline (Figure 6.1). A review (in preparation) that included
data of DRINK also suggests that the effect of removal of sugary drinks in overweight
or obese children may be even greater than in normal weight children.22 In addition,
Ebbeling et al.5,23 found that obese adolescents benefitted from the removal of sugary
drinks. Therefore my conclusion is that children with overweight or obesity especially
may benefit from the removal of sugar-sweetened beverages.
Future research
Further research should address the physiological mechanisms responsible for the
fattening effect of liquid calories. Future research should also focus on the effect that
can be attained by the replacement of sugar-sweetened by sugar-free beverages in a
real life setting, and whether the sugar or energy is to blame for the fattening effect
of liquid calories.
What physiological mechanisms are responsible for the lack of satiety obtained
from liquid sugars compared to solid foods?
Research suggests that solid calories induce higher levels of satiety compared to liquid
139
CHAPTER 6
Between-group difference in change in body
weight
0.5
0.0
kg
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
0
6
12
18
Follow-up (months)
Children with low or healthy BMI at baseline
Children with overweight or obese BMI at baseline
A
Standard deviation units
Between-group difference in change in BMI
z score
0.02
-0.02
-0.06
-0.10
-0.14
-0.18
-0.22
-0.26
0
6
12
18
Follow-up (months)
Children with low or healthy BMI at baseline
Children with overweight or obese BMI at baseline
B
Figure 6.1 Between group differences for body weight and BMI z score in the 477 completers
with low or healthy BMI at baseline versus overweight or obese BMI at baseline. Panel A
shows BMI z score with the between-group difference in the mean change from baseline (the
mean change in the BMI z score in the sugar-free group minus the mean change in the sugar
group), as a function of time. Panel B shows body weight with the between-group difference in
the mean change from baseline (the mean change in the body weight in the sugar-free group
minus the mean change in the sugar group), as a function of time.
140
GENERAL DISCUSSION
calories. A review involving over 40 preload studies revealed that a mean of 36% of an
energy challenge presented as a solid food or mixed meal was not compensated for
by a dietary adjustment.24 For semi-solid this was 79%, and for fluids 109%. The latter
number indicates that energy from beverages adds to, rather than substitutes for
other energy sources. However, this remains a controversial issue. Fluids have been
reported to possess stronger,25 similar,26 or weaker27-31 satiety properties compared
to solid foods. These seemingly discrepant findings have been attributed to
methodological differences across studies. However, I consider the evidence that
liquid calories indeed are less satiating than solid foods quite convincing.
The question remains which physiological mechanism is responsible for the
lower levels of satiety of liquid foods compared to solid foods. One hypothesis
postulates that eating rate or oro-sensory exposure time is responsible. Liquid foods
are consumed at an higher rate with a relatively short oro-sensory exposure time.32
Research has confirmed that prolonging the oro-sensory exposure time is associated
with earlier meal termination and/or higher levels of satiety.33 Also, in contrast to
solid foods, liquid foods run through the gastric tract without causing as much
stomach distension.34 This diminishes the cephalic phase responses, leading to
reduced satiety.35
The cephalic phase response refers to a set of physiological responses that
prepare the digestive system for the incoming flow of nutrients after ingestion.36 Part
of the cephalic phase response is the release of gastro-intestinal hormones such as
insulin, pancreatic polypeptide, leptin and ghrelin that induce satiety.37 Although
studies are not abundant, those which have been done suggest that cephalic phase
responses are indeed much smaller (absent) for liquids compared to solids. Teff et
al.38 showed that sweet-tasting liquids are no adequate stimuli for the emergence of
the cephalic phase insulin response. Also Ludwig speculates that liquid sugars reduce
the insulin spikes and thus diminish hunger.39 However, more research on the exact
physiological mechanisms is needed.
What is the effect of replacement of sugar-sweetened by sugar-free beverages in
the real life setting?
DRINK examined the effect of masked replacement of sugar-sweetened beverages by
sugar-free beverages on body weight in an experimental setting. The question
remains whether the effect of DRINK holds up when children replace their sugary
drinks by water because knowing what one drinks may lead to a different effect.
Ebbeling et al.23 found that overweight adolescents who knew that they consumed
141
CHAPTER 6
water and zero-calorie drinks for one year at home also gained less weight compared
to adolescents who continued to drink sugary beverages. This study suggests that
even when adolescents are aware of the sugar content, they gain less body weight
when consuming zero-calorie drinks.
Another question is whether the effect we found can be extrapolated to other
amounts of liquid sugar. Dutch children consume approximately 700 mL soft drink per
day.40 In DRINK we removed 250 mL of sugary drinks daily over an 18-months period.
This decreased weight gain by 1 kg. The question remains whether removal of 700 mL
of sugary drinks with also 10% sugar would then yield a difference in weight gain of
approximately 3 kg. However, instead of research on such extrapolations, I suggest
that future research examines the effect on weight gain of policies that aim at
reducing consumption of sugar-sweetened beverages and promote the intake of
water in (pre) school children.
Blaming the sugar in soft drinks for its fattening effect?
Another question is whether the same effect would occur with the removal of liquid
proteins or fats. Only a few studies have been done on this. The weak effects of liquid
calories on energy intake compensation were recently confirmed for liquid proteins,
fats and carbohydrates.31 This suggests that liquid calories in such substances
contribute to weight gain. However, the study of Maersk et al.41 found that milk
induced greater subjective fullness and less hunger than regular coke, but the caloric
compensation for both drinks was the same Therefore, in my opinion more research is
needed to compare feelings of satiety and energy intake after consumption of blinded
drinks with proteins, carbohydrates and fat.
Practical implications
Not only the DRINK trial, but also the synchronously published trial by Ebbeling et al.23
have provided compelling evidence that sugar-containing beverages are fattening.
This means that sufficient scientific support is available to underpin further
interventions designed to lower the consumption of sugar-sweetened beverages and
encourage the intake of sugar-free drinks. Sugar-sweetened beverages also include
fruit drinks. They are believed to be healthy. However, they contain a large number of
calories. Such interventions are especially important for the Netherlands with a
relatively high consumption of sugary beverages compared to other European
countries.40
142
GENERAL DISCUSSION
Community-based interventions to promote water intake
Interventions aimed at reducing the intake of sugar-sweetened drinks and promoting
the consumption of sugar-free beverages need to be incorporated into the current
obesogenic environment. The obesogenic environment has been described as the
‘sum of influences that the surroundings, opportunities, or conditions of life have on
promoting obesity in individuals or populations’.42 This means that obesity is a public
health problem rather than solely an individual matter.43 Environmental interventions
can be designed as community-based efforts. One such effort to prevent overweight
in school children was initiated in two small towns in France in 2000.44 This approach
turned out to be effective, and is now being extended to 200 towns in Europe, under
the name EPODE (Ensemble, Prevenons l’Obésité des Enfants “Together, let’s prevent
obesity in children”). The apparent success of such community-based interventions
suggests that a new approach to preventing and treating obesity should be
‘environment based’ involving and activating entire neighborhoods and communities.
This means that interventions aimed at reducing the intake of sugar-sweetened drinks
and promoting the consumption of sugar-free beverages should involve everyone in
the community from preschools to restaurants.
Consumption of sugar-free beverages
Interventions should aim at lowering the intake of sugar-sweetened beverages and
promoting the consumption of sugar-free beverages. Evidence is abundant that the
consumption of artificial-sweeteners is not harmful to human health.45,46 Since
artificial-sweeteners are considered to be safe, and provide hardly any calories they
are frequently used in food products, mainly in carbonated beverages.47 However,
one should be cautious with artificially-sweetened beverages because they have the
same impact on dental erosion as sugar-sweetened beverages due to the same levels
of acidity.48 Dental erosion has increased in the last decades49 and has been
associated not only with the consumption of fruit juice, (carbonated) soft drink,50,51
but also with the consumption of artificially-sweetened beverages. Due to the acidity
and not because of the sweetness, water should be promoted instead of artificiallysweetened beverages to diminish the likelihood of dental erosion.52
Stakeholders
As previously mentioned, community based interventions should be further initiated
to curb the consumption of sugar-sweetened beverages and promote intake of water.
143
CHAPTER 6
Community based interventions should involve as many stakeholders as possible
including government, industry, home, and (pre) school.
The government
The government could pursue different strategies to decrease the consumption of
sugar-sweetened drinks and increase the intake of sugar-free drinks in the form of for
example soda taxes, and regulations for portion sizes.
Soda tax
Soda tax could be a powerful weapon in efforts to reduce obesity, in the same way
that cigarette taxes have helped to curb smoking.53 Many US health experts are now
calling for higher taxes on soft drinks and various governments are considering or
have already introduced sugar-sweetened beverage taxes.54 Duffey et al.55 showed
that in the USA both body weight and the risk of diabetes decreased in communities
where soda and fast-food prices increased. It has been calculated that a penny-perounce tax on sugar-sweetened beverages could generate new tax revenues of $79
billion over 2010–2015.56 Such revenues could be used to support programs to reduce
childhood obesity.54 Food pricing strategies may be effective to enhance people’s
health.57 Recently, Waterlander et al.58 showed that discounting fruits and vegetables
is a promising strategy to increase fruits and vegetable purchases. However, evidence
for a beneficial tax strategy of sugar-sweetened beverages is limited. Studies need to
examine how consumers react to soda taxes, because there could be unintended
consequences, for example taxing soft drinks could result in consumers buying fewer
fruits and vegetables so they can still afford soft drinks. A study of Waterlander et al.59
(in preparation) showed that in adults a soda tax did not influence purchases in other
beverage categories, except for increased purchases of coffee and tea. However, it
could be argued that further policies on soda taxing should be postponed until such
taxing has been proven to be effective. Now that we know that sugar-sweetened
beverages contribute to an increase in body weight, such research should be
prioritized that takes different age groups into account.
Portion size limits
Governmental regulations may want to focus on limiting portion sizes. Portion sizes
have increased over the past decades60 and are specifically incorporated in major
promotional campaigns by the fast food industries. Larger servings provide more
calories, and previous research has shown that increased portion size leads to greater
consumption.61-64 Observational data suggest that larger portions of beverages in
particular, increase energy intake.65 This is problematic because people do not tend to
compensate for such increased energy intake.63 Smaller portions may therefore
144
GENERAL DISCUSSION
contribute to lower people’s energy intake.66,67 In children, scientific evidence is
lacking that sugary drinks in smaller portions reduce intake. However, in 2012 Mayor
Michael Bloomberg of New York City proposed a novel strategy to combat the rising
obesity crisis in the United States – banning the sale of sugar-sweetened beverages
larger than 16 ounces (0.5 liter). These initiatives should be seen as examples of how
to combat the intake of sugary beverages.
The food industry
There are various ways in which the food industry could take our findings into account
and convert these into new win-win strategies. It could lower the sugar content of
sweetened beverages or replace the sugar by artificial-sweeteners, and increase their
supply of water and sugar-free beverages. It could stop the marketing of sugarsweetened beverages to children and adolescents. According to research done by the
Rudd Center for Food Policy and Obesity at Yale, $4.2 billion was spent in 2009 by fast
food industries in advertising through media. The report also showed an increase in
television adverts to adolescents and teens in recent years. They reported a 21%
increase in adverts for preschoolers, 34% increase for children, and 39% increase for
teens since 2003.68 Globally different initiatives have been developed against such
marketing practices, for example in the United King,69 and in the Netherlands.70 In
Sweden fast food adverts aimed at children less than 12 years is banned.
Governmental regulations may be required to enforce such initiatives.
Increase of water availability at home and (pre) schools
Because children spend most of their days at school and in child care, (pre) schools
should be involved in teaching children to drink water. Interventions at all levels
should involve parents because when parents are involved, children are more likely to
drink water at home, and not only at school.71 Research has shown that parental
involvement is effective in school health promotion72,73 as they have a key influence
on the development and maintenance of children’s behaviors.74-76
Dietary preference and habits are initially established at the age of 0 to 4 years,
laying the foundation of eating habits later in life.77,78 Therefore preschool settings
should offer water instead of sugar-sweetened beverages. Research in preschools has
indicated various misperceptions on the part of caregivers about sugar-sweetened
beverages.71 Parents and teachers believed that sugar-sweetened beverages were
healthy due to mineral and vitamin content. Also, the vast majority of parents
overestimated the quality of their child's diet.79 These opinions resulted in a
perception of little need to change the drinking behavior of children. Therefore, an
intervention that promotes water should first explore the beliefs and attitudes of
145
CHAPTER 6
parents and teachers towards the drinking habits of their children. The intervention
may then provide preschool teachers with the means of encouraging them to offer
water. These could include the availability of water in classrooms. A previous study
indicated that water was available in most classrooms (84%), but was only accessible
by adults in half of those classrooms.80
Also elementary and secondary schools should promote water use. However,
elementary schools require a different approach to that of secondary schools given
the widespread prevalence of soft drink vending machines in secondary schools which
are prohibited in most elementary schools. Therefore secondary schools may consider
developing policies to reduce the availability of sugar-sweetened beverages in
vending machines.81 Elementary schools should involve parents because they provide
the drinks for their children to consume at school. Elementary schools may stop
asking parents to provide drinks, and supply water themselves instead.
Both elementary and secondary schools should make drinking water widely
accessible in schools. Previous intervention studies at both elementary and secondary
schools have shown that increasing access to drinking water in schools increased
water intake among students.82,83 However, the increase in water availability had no
effect on the sales of sugar-sweetened beverages.84 The latter could be explained by
the fact that children do not buy the sugary beverages at school but in nearby
supermarkets. Therefore solely placing water coolers at elementary and secondary
schools is insufficient. This should be part of a broader intervention.
All in all, I suggest that interventions to promote water start at an early age.
Water should become the habitual drink, both at home and at school. When children
in preschool settings are used to water, it will be easier for children to drink water
during the rest of their life. Children will not only benefit from water because it
replaces their habitual sugary drinks. Water consumption may also influence their
feeding pattern as a whole. A recent study has shown that that the key to getting
children to eat their greens may be to give them water with their meals. Researchers
found that children ate more raw vegetables such as carrots or peppers when they
had water with a meal than if they had soft drinks.85
Conclusions
DRINK showed that liquid sugars fulfill a unique role in causing weight gain. The
outcome of DRINK further supports evolving recommendations worldwide to
discourage sugar-sweetened drinks and promote non-caloric drinks. For the
146
GENERAL DISCUSSION
Netherlands in particular, these findings are of importance because the intake of
sugar-sweetened beverages is relatively high compared to that of other European
countries.40 Interventions should be set up in the form of community based
interventions.
Childhood obesity is a complex issue and sugar-sweetened beverages are only
one part of a larger picture. Such beverages are certainly not the only cause of the
obesity epidemic. However, removal of sugar-sweetened drinks may play a substantial
role in reducing weight gain in children. A recent editorial,86 and a review of the
DRINK trial87 both suggested that the results of DRINK provide a strong scientific base
for envisioned regulations to halt the epidemic of obesity by discouraging sugarsweetened beverages.
The research described in this thesis provides compelling evidence of the role of
sugary drinks in childhood obesity. “Remove sugary drinks, drink water instead” is a
simple message that can now be disseminated. One may speculate that there is no
other single food product whose elimination can contribute to a significant degree of
weight change. Sweet beverages are devoid of any essential nutritional ingredient and
one can therefore easily do without them. If we do not come to grips with the
growing obesity epidemic in children, this may be the first time in history where
children will have a shorter life expectancy than that of their parents.
147
CHAPTER 6
References
1. Sichieri R, Paula Trotte A, de Souza RA, Veiga GV. School randomised trial on
prevention of excessive weight gain by discouraging students from drinking sodas.
Public Health Nutr 2009;12:197-202.
2. Tordoff MG, Alleva AM. Effect of drinking soda sweetened with aspartame or
high-fructose corn syrup on food intake and body weight. Am J Clin Nutr 1990;51:963-9.
3. Tate DF, Turner-McGrievy G, Lyons E, 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;95:555-63.
4. James J, Thomas P, Cavan D, Kerr D. Preventing childhood obesity by reducing
consumption of carbonated drinks: cluster randomised controlled trial. BMJ
2004;328:1237-43.
5. Ebbeling CB, Feldman HA, Osganian SK, Chomitz VR, Ellenbogen SJ, Ludwig DS.
Effects of decreasing sugar-sweetened beverage consumption on body weight in
adolescents: a randomized, controlled pilot study. Pediatrics 2006;117:673-80.
6. Hrobjartsson A, Forfang E, Haahr MT, Als-Nielsen B, Brorson S. Blinded trials
taken to the test: an analysis of randomized clinical trials that report tests for the
success of blinding. Int J Epidemiol 2007;36:654-63.
7. de Ruyter JC, Olthof MR, Kuijper LD, Katan MB. Effect of sugar-sweetened
beverages on body weight in children: design and baseline characteristics of the
Double-blind, Randomized INtervention study in Kids. Contemp Clin Trials
2012;33:247-57.
8. Van Wymelbeke V, Beridot-Therond ME, de La Gueronniere V, Fantino M.
Influence of repeated consumption of beverages containing sucrose or intense
sweeteners on food intake. Eur J Clin Nutr 2004;58:154-61.
9. Rolls BJ, Kim S, Fedoroff IC. Effects of drinks sweetened with sucrose or
aspartame on hunger, thirst and food intake in men. Physiol Behav 1990;48:19-26.
10. Holt SH, Sandona N, Brand-Miller JC. The effects of sugar-free vs sugar-rich
beverages on feelings of fullness and subsequent food intake. Int J Food Sci Nutr
2000;51:59-71.
11. Canty DJ, Chan MM. Effects of consumption of caloric vs noncaloric sweet drinks
on indices of hunger and food consumption in normal adults. Am J Clin Nutr
1991;53:1159-64.
148
GENERAL DISCUSSION
12. Beridot-Therond ME, Arts I, Fantino M, De La Gueronniere V. Short-term effects
of the flavour of drinks on ingestive behaviours in man. Appetite 1998;31:67-81.
13. Rodin J. Comparative effects of fructose, aspartame, glucose, and water preloads
on calorie and macronutrient intake. Am J Clin Nutr 1990;51:428-35.
14. Trabulsi J, Troiano RP, Subar AF, et al. Precision of the doubly labeled water
method in a large-scale application: evaluation of a streamlined-dosing protocol in the
Observing Protein and Energy Nutrition (OPEN) study. Eur J Clin Nutr 2003;57:1370-7.
15. Schonbeck Y, Talma H, van Dommelen P, et al. Increase in prevalence of
overweight in dutch children and adolescents: a comparison of nationwide growth
studies in 1980, 1997 and 2009. PLoS One 2011;6:e27608.
16. Wishnofsky M. Caloric equivalents of gained or lost weight. Am J Clin Nutr
1958;6:542-6.
17. Wang YC, Gortmaker SL, Sobol AM, Kuntz KM. Estimating the energy gap among
US children: a counterfactual approach. Pediatrics 2006;118:e1721-33.
18. Antonetti VW. The equations governing weight change in human beings. Am J
Clin Nutr 1973;26:64-71.
19. Hall KD, Sacks G, Chandramohan D, et al. Quantification of the effect of energy
imbalance on bodyweight. Lancet;378:826-37.
20. De La Hunty A, Gibson S, Ashwell M. A review of the effectiveness of aspartame
in helping with weight control. Nutrition Bulletin 2006;31:115-28.
21. de Ruyter JC, Olthof MR, Seidell JC, Katan MB. A trial of sugar-free or sugarsweetened beverages and body weight in children. N Engl J Med 2012;367:1397-406.
22. Kaiser KA, Shikany JM, Keating KD, Allison DB. Will reducing sugar-sweetened
beverage consumption reduce obesity? Evidence supporting conjecture is strong, but
evidence when testing effect is weak. Obesity Reviews. In preperation
23. Ebbeling CB, Feldman HA, Chomitz VR, et al. A randomized trial of sugarsweetened beverages and adolescent body weight. N Engl J Med 2012;367:1407-16.
24. Mattes RD. Dietary compensation by humans for supplemental energy provided
as ethanol or carbohydrate in fluids. Physiology & Behavior 1995;59:179-87.
25. Kissileff HR. Effects of physical state (liquid-solid) of foods on food intake:
procedural and substantive contributions. Am J Clin Nutr 1985;42:956-65.
26. Tournier A, Louis-Sylvestre J. Effect of the physical state of a food on subsequent
intake in human subjects. Appetite 1991;16:17-24.
27. Haber GB, Heaton KW, Murphy D, Burroughs LF. Depletion and disruption of
dietary fibre. Effects on satiety, plasma-glucose, and serum-insulin. Lancet
1977;2:679-82.
149
CHAPTER 6
28. Bolton RP, Heaton KW, Burroughs LF. The role of dietary fiber in satiety, glucose,
and insulin: studies with fruit and fruit juice. Am J Clin Nutr 1981;34:211-7.
29. Hulshof T, De Graaf C, Weststrate JA. The effects of preloads varying in physical
state and fat content on satiety and energy intake. Appetite 1993;21:273-86.
30. DiMeglio DP, Mattes RD. Liquid versus solid carbohydrate: effects on food intake
and body weight. Int J Obes Relat Metab Disord 2000;24:794-800.
31. Mourao DM, Bressan J, Campbell WW, Mattes RD. Effects of food form on
appetite and energy intake in lean and obese young adults. Int J Obes (Lond)
2007;31:1688-95.
32. Viskaal-van Dongen M, Kok FJ, de Graaf C. Eating rate of commonly consumed
foods promotes food and energy intake. Appetite 2011;56:25-31.
33. Zijlstra N, Mars M, de Wijk RA, Westerterp-Plantenga MS, de Graaf C. The effect
of viscosity on ad libitum food intake. Int J Obes (Lond) 2008;32:676-83.
34. Lewis, FT. The form of the stomach in human embryos with notes upon the
nomenclature of the stomach. Am J Anat 1912;13:477-503.
35. de Graaf C. Why liquid energy results in overconsumption. Proc Nutr Soc
2011;70:162-70.
36. Zafra MA, Molina F, Puerto A. The neural/cephalic phase reflexes in the
physiology of nutrition. Neurosci Biobehav Rev 2006;30:1032-44.
37. Powley TL, Berthoud HR. Diet and cephalic phase insulin responses. Am J Clin
Nutr 1985;42:991-1002.
38. Teff KL, Devine J, Engelman K. Sweet taste: effect on cephalic phase insulin
release in men. Physiol Behav 1995;57:1089-95.
39. Ludwig DS. The glycemic index: physiological mechanisms relating to obesity,
diabetes, and cardiovascular disease. JAMA 2002;287:2414-23.
40. Brug J, van Stralen MM, Te Velde SJ, et al. Differences in weight status and
energy-balance related behaviors among schoolchildren across Europe: the ENERGYproject. PLoS One 2012;7:e34742.
41. Maersk M, Belza A, Holst JJ, et al. Satiety scores and satiety hormone response
after sucrose-sweetened soft drink compared with isocaloric semi-skimmed milk and
with non-caloric soft drink: a controlled trial. Eur J Clin Nutr 2012;66:523-9.
42. Swinburn B, Egger G, Raza F. Dissecting obesogenic environments: the
development and application of a framework for identifying and prioritizing
environmental interventions for obesity. Prev Med 1999;29:563-70.
43. Veerman JL, Barendregt JJ, van Beeck EF, Seidell JC, Mackenbach JP. Stemming
the obesity epidemic: a tantalizing prospect. Obesity (Silver Spring) 2007;15:2365-70.
150
GENERAL DISCUSSION
44. Borys JM, Le Bodo Y, Jebb SA, et al. EPODE approach for childhood obesity
prevention: methods, progress and international development. Obes Rev
2012;13:299-315.
45. Baird IM, Shephard NW, Merritt RJ, Hildick-Smith G. Repeated dose study of
sucralose tolerance in human subjects. Food Chem Toxicol 2000;38 Suppl 2:S123-9.
46. NTP toxicology studies of acesulfame potassium (CAS No. 55589-62-3) in
genetically modified (FVB Tg.AC Hemizygous) mice and carcinogenicity studies of
acesulfame potassium in genetically modified [B6.129-Trp53(tm1Brd) (N5)
Haploinsufficient] mice (feed studies)mice. Natl Toxicol Program Genet Modif Model
Rep. 2005;2:1-113.
47. Mattes RD, Popkin BM. Nonnutritive sweetener consumption in humans: effects
on appetite and food intake and their putative mechanisms. Am J Clin Nutr 2009;89:1-14.
48. Lussi A, Jaeggi T. Chemical factors. Monogr Oral Sci 2006;20:77-87.
49. Nunn JH, Gordon PH, Morris AJ, Pine CM, Walker A. Dental erosion -- changing
prevalence? A review of British National childrens' surveys. Int J Paediatr Dent
2003;13:98-105.
50. Dugmore CR, Rock WP. A multifactorial analysis of factors associated with dental
erosion. Br Dent J 2004;196:283-6.
51. Hasselkvist A, Johansson A, Johansson AK. Dental erosion and soft drink
consumption in Swedish children and adolescents and the development of a
simplified erosion partial recording system. Swed Dent J 2010;34:187-95.
52. Gambon DL, Brand HS, Veerman EC. Dental erosion in the 21st century: what is
happening to nutritional habits and lifestyle in our society? Br Dent J 2012;213:55-7.
53. Chaloupka FJ, Straif K, Leon ME. Effectiveness of tax and price policies in tobacco
control. Tob Control 2011;20:235-8.
54. Brownell KD, Farley T, Willett WC, et al. The public health and economic benefits
of taxing sugar-sweetened beverages. N Engl J Med 2009;361:1599-605.
55. Duffey KJ, Gordon-Larsen P, Shikany JM, Guilkey D, Jacobs DR, Jr., Popkin BM.
Food price and diet and health outcomes: 20 years of the CARDIA Study. Arch Intern
Med 2010;170:420-6.
56. Andreyeva T, Chaloupka FJ, Brownell KD. Estimating the potential of taxes on
sugar-sweetened beverages to reduce consumption and generate revenue. Prev Med
2011;52:413-6.
57. McCarthy M. The economics of obesity. Lancet 2004;364:2169-70.
58. Waterlander WE, de Boer MR, Schuit AJ, Seidell JC, Steenhuis IH. Price discounts
significantly enhance fruit and vegetable purchases when combined with nutrition
education: a randomized controlled supermarket trial. Am J Clin Nutr 2013;97:886-95.
151
CHAPTER 6
59. Waterlander WE, Ni Mhurchu C, Steenhuis IHM. Effects of a tax increase on
sugar containing beverages: results from a randomized controlled trial. Under review.
60. Nielsen SJ, Popkin BM. Changes in beverage intake between 1977 and 2001. Am
J Prev Med 2004;27:205-10.
61. Wansink B, Kim J. Bad popcorn in big buckets: portion size can influence intake
as much as taste. J Nutr Educ Behav 2005;37:242-5.
62. Rolls BJ, Morris EL, Roe LS. Portion size of food affects energy intake in normalweight and overweight men and women. Am J Clin Nutr 2002;76:1207-13.
63. Rolls BJ, Roe LS, Kral TV, Meengs JS, Wall DE. Increasing the portion size of a
packaged snack increases energy intake in men and women. Appetite 2004;42:63-9.
64. Wansink B, Payne CR, Chandon P. Internal and external cues of meal cessation:
the French paradox redux? Obesity (Silver Spring) 2007;15:2920-4.
65. Matthiessen J, Fagt S, Biltoft-Jensen A, Beck AM, Ovesen L. Size makes a
difference. Public Health Nutr 2003;6:65-72.
66. Freedman MR, Brochado C. Reducing portion size reduces food intake and plate
waste. Obesity (Silver Spring) 2010;18:1864-6.
67. Rolls BJ, Roe LS, Meengs JS. Reductions in portion size and energy density of
foods are additive and lead to sustained decreases in energy intake. Am J Clin Nutr
2006;83:11-7.
68. Harris JL, Schwartz MB, Brownell KD et al. Fast Food f.a.c.t.s.: Food Advertising
to Children and Teen Score. http://www.fastfoodmarketing.org/. Accessed at April 27,
2013.
69. House of Commons. Children's Food Bill.
http://www.publications.parliament.uk. Accessed at April 27, 2013.
70. De Vries HA. Over de noodzaak van een gezond aanbod. Kindermarketing:
Onverantwoord en ongereguleerd (English translation: About the need for a healthy
supply. Children Marketing: irresponsible and unregulated. http://www.foodwatch.nl.
Accessed at April 27, 2013
71. De Craemer M, De Decker E, De Bourdeaudhuij I, et al. Physical activity and
beverage consumption in preschoolers: focus groups with parents and teachers. BMC
Public Health 2013;13:278.
72. Peters LW, Kok G, Ten Dam GT, Buijs GJ, Paulussen TG. Effective elements of
school health promotion across behavioral domains: a systematic review of reviews.
BMC Public Health 2009;9:182.
73. Cong Z, Feng D, Liu Y, Esperat MC. Sedentary behaviors among Hispanic children:
influences of parental support in a school intervention program. Am J Health Promot
2012;26:270-80.
152
GENERAL DISCUSSION
74. Vereecken CA, Keukelier E, Maes L. Influence of mother's educational level on
food parenting practices and food habits of young children. Appetite 2004;43:93-103.
75. Grimm GC, Harnack L, Story M. Factors associated with soft drink consumption
in school-aged children. J Am Diet Assoc 2004;104:1244-9.
76. Brotman LM, Dawson-McClure S, Huang KY, et al. Early childhood family
intervention and long-term obesity prevention among high-risk minority youth.
Pediatrics 2012;129:e621-8.
77. Birch LL. Development of food preferences. Annu Rev Nutr 1999;19:41-62.
78. Birch L, Savage JS, Ventura A. Influences on the Development of Children's Eating
Behaviours: From Infancy to Adolescence. Can J Diet Pract Res 2007;68:s1-s56.
79. Kourlaba G, Kondaki K, Grammatikaki E, Roma-Giannikou E, Manios Y. Diet
quality of preschool children and maternal perceptions/misperceptions: the GENESIS
study. Public Health 2009;123:738-42.
80. Middleton AE, Henderson KE, Schwartz MB. From policy to practice:
imple