Open access Research

BMJ Open: first published as 10.1136/bmjopen-2017-021387 on 18 September 2018. Downloaded from http://bmjopen.bmj.com/ on 29 November 2018 by guest. Protected by copyright.
Evaluation of the nutrient content of
yogurts: a comprehensive survey of
yogurt products in the major
UK supermarkets
J Bernadette Moore,1,2 Annabelle Horti,1 Barbara A Fielding2

To cite: Moore JB, Horti A, Abstract

Fielding BA. Evaluation of the Strengths and limitations of this study
Objectives  To comprehensively survey the sugar and
nutrient content of yogurts: nutrient contents of yogurt products available in UK
a comprehensive survey of ►► This was a comprehensive market survey that anal-
supermarkets, in particular those marketed to children.
yogurt products in the major ysed the nutrient contents of 921 supermarket prod-
Design  A cross-sectional survey of yogurt products
UK supermarkets. BMJ Open ucts identified by the search term yogurt/yoghurt in
2018;8:e021387. doi:10.1136/ available in the UK’s supermarkets in November 2016.
five major UK supermarkets, representing 75% of
bmjopen-2017-021387 Methods  Data were collected from five major online
UK grocery market share, online in November 2016.
UK supermarkets and a process flow strategy was used
►► Prepublication history for ►► A systematic process flow strategy was determined
to place yogurts into eight categories: children’s, dairy
this paper is available online. a priori and used for categorising products.
alternatives, dessert, drinks, fruit, flavoured, natural/Greek
To view these files please visit ►► The energy and nutrient contents of products within
style and organic. A comprehensive database of product
the journal online (http://​dx.​doi.​ categories are individually presented readily permit-
org/​10.​1136/​bmjopen-​2017-​ information for 921 unique products was created and
ting differences and ranges between categories to
021387). analysed.
be seen.
Results  The total sugar, fat, protein, calcium and energy
►► Products were identified by supermarket categori-
Received 23 December 2017 contents were highly variable across categories, and
Revised 28 May 2018
sation and some products included in this market
the ranges were extremely broad. Although lower than
Accepted 20 June 2018 survey for completeness did not contain yogurt cul-
the dessert category, the medians (range) of the total
tures (eg, fromage frais, desserts), which may have
sugar content of children’s (10.8 g/100 g (4.8–14.5)),
skewed results.
fruit (11.9 g/100 g (4.6–21.3)), flavoured (12.0 g/100 g
►► The nutrient information used here came from the
(0.1–18.8)) and organic (13.1 g/100 g (3.8–16.9)) yogurt
manufacturers and supermarkets online and may
products were all well above 10 g/100 g, and represented
not have been accurate or up to date; and the sur-
>45% of total energy. Only two out of 101 children’s
vey did not include less well-known yogurt brands,
yogurt and fromage frais products surveyed qualified
produced or sold by small-scale local food manufac-
as low sugar (≤5 g/100 g). Natural/Greek yogurts had
turers, or yogurts sold in dessert shops.
dramatically lower sugar contents (5.0 g/100 g (1.6, 9.5),
largely lactose) than all other categories. While low-fat
(<3 g/100 g) products had less sugar and energy than
higher fat yogurts, nonetheless 55% (285 of 518 low-fat products.1 Yogurt is the product of milk
yogurts) contained between 10 and 20 g sugar/100 g. fermented with the lactic cultures Streptococcus
Within the children’s category, fromage frais had higher thermophiles and Lactobacillus delbrueckii subsp
protein (5.3 g/100 g (3.3, 8.6) vs 3.2 (2.8, 7.1); p<0.0001) bulgaricus, which can be enhanced with other
and calcium contents (150 mg/100 g (90, 240) vs probiotic cultures such as Lactobacillus acidoph-
© Author(s) (or their 130.5 mg/100 g (114, 258); p=0.0015) than yogurts. ilus and Bifidobacterium bifidus.2 Consumption
employer(s)) 2018. Re-use Conclusions  While there is good evidence that yogurt
permitted under CC BY-NC. No of fermented dairy products has long been
can be beneficial to health, products on the market vary
commercial re-use. See rights considered to be beneficial to digestive and
widely in total sugars. Fewer than 9%, and only 2% of
and permissions. Published by
the children’s, products surveyed were low enough in overall health.3 The beneficial probiotic
BMJ. and immune regulatory effects of yogurts
1 sugar to earn ‘green’ in UK front of the pack labelling.
School of Food Science and underpin their recommendation as a healthy
Reformulation for the reduction of free sugars in yogurts is
Nutrition, University of Leeds,
Leeds, UK warranted. food for babies and children.4 Multiple regu-
2
Department of Nutritional latory bodies, including the European Food
Sciences, Faculty of Health and Safety Authority, have approved health claims
Medical Sciences, University of Introduction   related to yogurt consumption and reduc-
Surrey, Surrey, UK The association between dairy foods and tion in symptoms caused by lactose maldi-
Correspondence to disease risk has often been contradictory, gestion.5 In addition to probiotics, yogurt is
Dr J Bernadette Moore; likely due to inherent diversity in the nutrient a good source of protein, calcium, iodine and
​J.​B.​Moore@​leeds.​ac.​uk contents and food matrices of different dairy vitamin B12, and its consumption has been

Moore JB, et al. BMJ Open 2018;8:e021387. doi:10.1136/bmjopen-2017-021387 1

Open access

BMJ Open: first published as 10.1136/bmjopen-2017-021387 on 18 September 2018. Downloaded from http://bmjopen.bmj.com/ on 29 November 2018 by guest. Protected by copyright.
associated with lower risk of obesity and cardiometabolic
Table 1  Definitions
risk in both children and adults.6 7
Evidence is accumulating that frequent yogurt Term Definition
consumption may be associated with healthier meta- Sugars* Conventionally describes chemically the
bolic profiles in both children and adults.8 9 In adults, monosaccharides (glucose, fructose,
increased yogurt consumption has been associated with galactose) and disaccharides (sucrose,
lower levels of circulating triglycerides, glucose and lower lactose†, maltose).
systolic blood pressure9; and several recent meta-analyses Total sugars Currently required for UK nutrition label.
have demonstrated that increased yogurt consumption Includes sugars occurring naturally in foods
is inversely associated with the risk of developing type 2 and beverages and those added during
processing and preparation.
diabetes (T2D).10–13 Notably, across three large American
cohort studies (the Health Professionals Follow-Up Study Free sugars ‘All monosaccharides and disaccharides
and the Nurses’ Health Studies I and II) in >190 000 adults added to foods by the manufacturer, cook
or consumer, plus sugars naturally present
and >30 years of follow-up, while there was no effect of
in honey, syrups and unsweetened fruit
dairy on incident T2D, yogurt intake specifically was juices. Under this definition lactose† when
inversely associated with T2D risk across the three cohorts naturally present in milk and milk products
with a pooled HR of 0.83 (0.75, 0.92) for one serving/ is excluded.’36 Sugars present in intact fruits
day.12 Although confounders were statistically accounted and vegetables are also excluded.
for, an acknowledged limitation to these epidemiology Added sugars ‘Syrups and other caloric sweeteners used
studies is the evidence that yogurt is a general marker as a sweetener in other food products.
of healthy dietary habits.9 14 Data are more equivocal Naturally occurring sugars such as those
regarding yogurt consumption and cardiovascular disease in fruit or milk are not added sugars.’61 In
risk, although one study suggested a possible lowered addition, excludes sugars in juiced or pureed
risk at higher levels of yogurt intakes (>200 g/day),15 a fruits and vegetables that are included in
definition of free sugars. Will be a required
more recent and comprehensive meta-analysis showed
subline under ‘total sugars’ for US food
no benefit.16 High-quality and adequately powered labels from 2020.46
randomised controlled trials are lacking however.
Prospective cohort studies have, in general, shown Adapted with permission from Moore and Fielding [34].
*Examples of sugars commonly found as ingredients: sucrose,
inverse associations between yogurt intakes and changes
fructose, glucose, dextrose, maltose, lactose, trehalose, brown
in waist circumference, weight and risk of overweight or sugar, turbinado sugar, demerara sugar, raw sugar, cane sugar, fruit
obesity; with some inconsistencies between studies.17 18 sugar, invert sugar, corn sweetener, corn syrup, high-fructose corn
More recently, in an elderly Italian population cohort syrup, malt syrup, glucose syrup, glucose-fructose syrup, fructose-
of >4000 at high cardiovascular risk, consumption of glucose syrup, honey, molasses, date syrup, agave syrup.
†Lactose is often called ‘milk sugar’ because 100% of ‘total
whole-fat yogurt (but not total yogurt) was associated
sugars’ in milk are lactose. In natural/Greek yogurt ~80% of the
with changes in waist circumference and higher proba- sugar is lactose, with the remainder being galactose generated
bility for reversion of abdominal obesity and lower risk of from lactose fermentation.62
diabetes.19 20 This is in line with a previous systematic review
of observational studies on the relationship between dairy
fat, obesity and cardiometabolic disease; in 11 of 16 studies In the UK, on average, children consume more
included in the review, high-fat dairy intake was inversely yogurt than adults, and children under 3 years of age
associated with measures of adiposity.21 While dietary have the highest intakes.27 Yogurt contributes a signifi-
guidelines vary by region, most countries make dietary cant percentage of the daily Reference Nutrient Intake
recommendations for the consumption of dairy prod- of key nutrients to babies and children up to 10 years
ucts because of the strong evidence for the role of dairy of age, providing, for example: 10%–19%  calcium,
products in meeting nutrient intake requirements.22 In 11%–20% phosphorus, 10%–21% riboflavin and
both the USA and UK, current dietary guidelines recom- 17%–54% vitamin B12.27 Yogurts are often recommended
mend low-fat and low-sugar dairy products because of to be part of children’s diets due to their high calcium
obesity-related concerns; however, a growing number of content and its positive effect on bone development.28
recent studies suggest that high-fat dairy consumption is Calcium also has a positive effect on teeth and high
associated with a lower risk of obesity and diabetes.21 23–25 intakes of milk and yogurt products in multiple studies
Large cohort studies in Sweden (n=1782 men 40–60 years have been associated with reduced tooth erosion.29
old)23 and the USA (n=18 438 women ≥45 years old in However, although there is good evidence to suggest that
the Women's Health Study)24 have found high-fat dairy yogurt can be beneficial to health, products on the market
to be protective against developing central adiposity and may vary widely in sugar content and yogurt marketed
becoming overweight or obese at follow-up. These data specifically to children may be higher in free or added
have fed into ongoing wider debate regarding dietary sugars (see table 1 for definitions of free, added and
guidance related to fat, refined carbohydrates and cardio- total sugars).30 31 Dairy is a significant contributor to the
vascular disease risk.26 intakes of free sugars by children and adults.32 Diets high

2 Moore JB, et al. BMJ Open 2018;8:e021387. doi:10.1136/bmjopen-2017-021387

 Open access

BMJ Open: first published as 10.1136/bmjopen-2017-021387 on 18 September 2018. Downloaded from http://bmjopen.bmj.com/ on 29 November 2018 by guest. Protected by copyright.
in free sugars are now unequivocally linked to obesity panna cotta, chilled soya desserts, quark, egg custards’,39
and dental caries, prompting the WHO and other regu- we chose to include these products in the analyses here.
latory bodies in updating dietary guidelines to strongly Therefore, the dessert category contains both yogurt-
advocate for restricting free sugar intakes to less than based and other products (eg, jellies and puddings, ie,
10% of total energy.33 34 With an alarming 58% of women chocolate mousse, crème caramels) that contain no
and 68% of men along with one in three of UK children yogurt or fromage frais cultures. However, non-yo-
aged 10–11 years overweight or obese in 2015,35 the UK’s gurt (eg, high-protein) drinks were removed from the
guidelines more stringently recommended the restric- nutrient analyses. Soya-based yogurts were placed in the
tion of free sugars to less than 5% of total daily energy.36 'dairy alternative’ category, whereas Greek-style yogurts
As part of a plan to combat childhood obesity, in April with added honey or other sweeteners were placed in
2018 the UK government implemented an industry soft the ‘flavoured’ category. ‘Fruit’ was defined liberally, for
drinks levy and commissioned a structured programme example, many products were made with either curd or
of monitored sugar reduction as part of wider refor- purees and for lemon products in particular, often with
mulation tackling calories, salt and saturated fat.37 The juice. In the case of natural yogurts, if organic, they went
initial focus was on the top nine food categories (after in organic category. In scrutinising the children’s cate-
soft drinks and fruit juices and smoothies) that contribute gory (products included were defined as ‘children’s’
the most to children’s sugar intakes. These are: chocolate, if either the supermarket or product itself defined it as
confectionery, biscuits, breakfast cereals, cakes, morning such, or if spokescharacters, celebrities, cartoons, toy
goods (such as croissants, buns and waffles), ice cream, giveaways, games or kids’ clubs were incorporated into
yogurt and sweet spreads/sauces. yogurt was one of the brand image), the ingredient lists were used to evaluate
products identified and highlighted for a 20% reduction presence or absence of yogurt or fromage frais cultures.
of sugar by 2020, with guidelines given for energy per For the classification of low fat and low sugar, the cut-offs
portion size of 120 kcal sales weighted average; 175 kcal defined by European Union (EU) regulations2 currently
maximum per portion and an allowance made for lactose used for voluntary, front-of-pack, traffic light coloured
(3.8 g/100 g).37 labels in the UK40 were used. In this scheme, green desig-
In this context then, the aims of this work were to nates low, amber designates medium and red designates
perform a comprehensive survey of yogurt products high levels of reference intakes. Low fat is defined as ≤3 g
within the major UK supermarkets, in order to identify of fat/100 g or ≤1.5 g/100 mL for drinks; low sugar is
products marketed at children and to evaluate their
nutrient contents, in particular sugar, compared with
other categories.

Experimental methods
Data collection
Data were collected from five major UK online super-
markets (Asda, Morrisons, Sainsbury’s, Tesco and Wait-
rose) that account for 75% of the UK grocery market
share.38 Websites were searched from 7 October 2016 to
16 November 2016 using ‘yogurt’ or ‘yoghurt’ as a search
term (in four of five supermarkets the same number of
products was returned). After considering the product
groupings commonly used by online supermarkets, eight
categories (children’s, dairy alternatives, dessert, drinks,
flavoured, fruit, natural/Greek, organic) and a systematic
process flow strategy for product placement (figure 1)
were decided on a priori. These product groupings, and
those typically used by supermarkets for both inventory
and marketing reasons, go beyond the standard broad
subgroupings used in dietary surveys such as the UK
National Dietary and Nutrition Survey (NDNS), where
the main food group ‘15: yogurt, fromage frais and other
dairy desserts’ has only three broad subsidiary groups of
either: ‘yogurt’; ‘manufactured fromage frais and other Figure 1  Process flow diagram of category decisions. Data
dairy desserts’; or ‘homemade fromage frais and other were collected using yogurt as a search term within the UK’s
dairy desserts’.39 As dairy desserts for the NDNS include top five online supermarkets between 7 October 2016 and
‘chocolate and fruit cream desserts, mousse, milk jelly, 16 November 2016. Products were classified into different
junket, buttermilk desserts, fruit fools, crème caramel, categories as shown.

Moore JB, et al. BMJ Open 2018;8:e021387. doi:10.1136/bmjopen-2017-021387 3

Open access

BMJ Open: first published as 10.1136/bmjopen-2017-021387 on 18 September 2018. Downloaded from http://bmjopen.bmj.com/ on 29 November 2018 by guest. Protected by copyright.

Figure 2  Nutrient and energy contents of UK yogurt products across categories. (A) Sugar. (B) Fat. (C) Protein. (D) Calcium.
(E) Energy. (F) Energy/serving. Data were tested for normality and analysed using the Kruskal-Wallis and Dunn’s multiple
comparison tests; categories with unlike letters were significantly different. Median is indicated by black line. Dashed lines
indicate thresholds defined by European Union (EU) regulations2 for nutrition claims for low sugar (A) and low fat (B).

defined by a maximum of 5 g total sugars/100 g. For label- energy from sugars and is much higher than the recent
ling, high fat and high sugar (red) are defined at >17.5 g UK and WHO targets of 5%–10% of energy from free
fat/100 g and >22.5 g sugar/100 g, respectively. Notably sugars (25–50 g sugar/day in a 2000 kcal diet).2 40
for sugar, the current EU reference intakes are for 90 g Data were screened for duplicates and a non-redun-
sugar in a 2000 kcal diet; this represents 18% of total dant database of product information was created that

4 Moore JB, et al. BMJ Open 2018;8:e021387. doi:10.1136/bmjopen-2017-021387

 Open access

BMJ Open: first published as 10.1136/bmjopen-2017-021387 on 18 September 2018. Downloaded from http://bmjopen.bmj.com/ on 29 November 2018 by guest. Protected by copyright.
Table 2  Macronutrients across yogurt categories
Fat Carbohydrates (total sugars) Protein
n Median Range Median Range Median Range
(% energy)
Children’s 101 26.2* 19.9–40.8 52.0* (45.5) 25.4–57.9 20.0* 13.1–40.5
Dairy
alternatives 38 30.2* 21.9–86.3 50.8*† (48.4) 7.1–63.5 18.8* 2.4–40.0
Dessert 161 32.6* 0.0–63.1 55.0‡ (46.3) 29.0–100 10.8† 0.0–34.9
Drinks 70 17.1† 0.0–54.8 62.4‡ (52.5) 25.6–95.2 15.8† 7.9–40.0
Flavoured 79 31.7* 0.0–53.5 52.2*†  (45.8) 26.4–69.6 14.7† 0.0–70.0
Fruit 317 16.6† 0.0–53.5 56.8‡ (52.8) 25.5–78.1 17.7† 7.6–61.5
Natural/Greek 61 25.9* 0.0–75.0 34.7† (30.4) 11.3–61.3 32.3‡ 11.4–72.3
Organic 71 33.4* 0.0–69.8 48.5*† (46.7) 16.4–73.5 17.5* 0.3–56.3
(g/100  g)
Children’s 101 2.8* 1.9–5.7 12.3* (10.8) 4.9–25.0 5.3* 2.8–8.6
Dairy
alternatives 38 2.5*† 1.9–21.0 9.5* (9.2) 1.0–16.2 3.6† 0.6–5.2
Dessert 161 5.2† 0.0–26.7 19.6† (16.4) 6.3–54.9 4.0†‡ 2.0–9.5
Drinks 70 1.5‡ 0.0–3.0 11.2* (9.1) 2.3–18.4 2.7† 1.3–5.9
Flavoured 79 3.6*† 0.0–9.6 13.0* (12.0) 3.7–19.0 4.2*‡§ 0.4–9.8
Fruit 317 1.6§ 0.0–8.9 12.9* (11.9) 4.8–22.4 4.2§ 2.1–10.0
Natural/Greek 61 1.7*†§ 0.0–10.2 5.5‡ (5.0) 3.6–9.5 5.4* 2.2–11.0
Organic 71 3.9*† 0.0–10.1 13.3* (13.1) 4.8–22.7 4.5*§ 3.0–7.6
*†‡§Median values within a column with unlike superscript symbols were significantly different (p<0.0001) by Kruskal-Wallis and Dunn's
multiple comparison tests.

included: nutrient information, serving size, size of pack, Patient and public involvement
claims on pack and ingredients. Nutrient information was Patients and the public were not involved in this research.
confirmed from the brand’s own website where possible
and a subset was examined during in-store visits. Informa-
tion on macronutrients, including energy, fat, saturated Results
fat, carbohydrates, sugar, fibre and protein, was collected, Nine hundred and twenty-one products identified online
as well as any information on micronutrients. Data for during the period of the survey were initially included
macronutrients have been expressed as g/100 g product in the analysis. At the time of survey, although national
as well as % energy, as both are needed to evaluate its branded products dominated the products available
nutritional merits; for example, a yogurt with a lower (n=648 national vs n=273 own brand products), only 65
energy content per 100 g may have a higher % energy products were available in all five supermarkets. Following
from sugar than a product with a higher energy content. the process flow strategy (figure 1) removed 23 non-yo-
All data were independently double-checked and 5% of gurt, for example, high protein type, beverages from the
all entries were randomly selected and verified. analyses and the remaining 898 products were classed
as either: children’s (n=101), dairy alternatives (n=38),
Data analysis dessert (n=161), drinks (n=70), flavoured (n=79), fruit
Pivot tables in Excel were used for building and manip- (n=317), natural/Greek (n=61) or organic (n=71).
ulating the product database and statistical analyses In assessing nutrient contents across the surveyed prod-
were done utilising GraphPad Prism V.7.0 c. Normality ucts, the sugar content varied enormously both within
was examined using the D’Agostino-Pearson omnibus and across our product categories (figure 2A). With the
normality test and comparisons across all categories were exception of the natural/Greek category, the median
made using the non-parametric Kruskal-Wallis test with total sugar contents of all categories were well above the
Dunn’s multiple comparisons. For comparisons of two 5 g/100 g maximal threshold considered for a low-sugar
categories, again in not observing normal distributions, nutrition label claim.2 40 Products within the dessert
the non-parametric, two-tailed Mann-Whitney tests were category, unsurprisingly, had the highest median and
applied. broadest (range) of total sugar at 16.4 g/100 g (1.5–32.6)

Moore JB, et al. BMJ Open 2018;8:e021387. doi:10.1136/bmjopen-2017-021387 5

Open access

BMJ Open: first published as 10.1136/bmjopen-2017-021387 on 18 September 2018. Downloaded from http://bmjopen.bmj.com/ on 29 November 2018 by guest. Protected by copyright.
threshold with medians for fat of 3.6 g/100 g (0, 9.6) and
3.9 g/100 g (0, 10.1), respectively, the dessert category
contained the highest median amount of fat and had the
broadest range at 5.2 g/100 g (0–26.7) (figure 2B). The
drinks category had the lowest median fat contents at
1.5 g/100 g (0, 3.0), but the fruit and natural/Greek cate-
gories were also relatively low with medians of 1.6 (0, 8.9)
and 1.7 (0, 10.1) g/100 g, respectively (figure 2B).
The natural/Greek and the children’s categories
had the highest median protein contents at 5.4 (2.2,
11.0) and 5.3 (2.8, 8.6) g/100 g yogurt (figure 2C).
The higher protein median for the children’s category
was influenced, as expanded on below, to a significant
degree by a large percentage of higher protein fromage
frais products. The dessert, fruit, flavoured and organic
categories had very similar median protein contents
at 4.0–4.5 g/100 g, whereas drinks and dairy alternatives
contained the least amount of protein at 2.7 (1.3, 5.9)
and 3.6 (0.6, 5.2) g/100 g (figure 2C). Calcium values
were less frequently reported, but median values between
yogurt categories were broadly similar ranging from 116
to 150 mg/100 g (figure 2D). Mimicking protein, the
natural/Greek and the children’s categories had the
highest calcium contents. The dessert category, again not
surprisingly, contained significantly more energy/100 g
(figure 2E) and more energy/serving (figure 2F) than any
other category. Although the children’s category had the
least energy/serving (figure 2F), this was clearly due to
smaller serving sizes. When expressed as kcal/100 g, the
Figure 3  Macronutrients compared across low-fat median energy of children’s yogurts was similar to prod-
(≤3 g/100 g; n=530) and higher fat (>3 g/100 g; n=383) ucts in the dairy alternative, flavoured, fruit and organic
products. (A) Sugar. (B) Fat. (C) Protein. (D) Energy. Data were categories (ranging from 79 to 100 g/100 g), and these
tested for normality and analysed using the Mann-Whitney
were significantly higher than the median (65 g/100 g) of
test. Median is indicated by black line. ****P<0.0001.
the natural/Greek category (figure 2E).
In examining the macronutrient content (as % energy)
(figure 2A). However, the children’s, flavoured, fruit and of yogurts across categories, it was clear that the majority
organic categories all had relatively high, and similar, of carbohydrates in yogurt products are derived from
median total sugar contents ranging from 10.8 g/100 g sugars (table 2). Carbohydrate content ranged from the
(children’s) to 13.1 g/100 g (organic). As the natural/ extremes of 34.7% (natural/Greek) to 62.4% (drinks)
Greek category by definition contained no added sweet- of energy content, but all other categories were tightly
eners, it had the lowest median amount of total sugars bunched at 48.5%–56.7%  carbohydrate content. The
per product (5 g/100 g (1.6, 9.5)). These sugars will natural/Greek category was significantly higher in
be ~80% lactose (table 1), but notably the range observed protein (32.3%) than any other category, while the dessert
here suggests many products will have higher than the category had the least amount of protein (10.8%) and
allowance of 3.8 g/100 g specified for lactose in the UK highest percentage of fat (32.6%; table 2). Although fruit
sugar reduction programme guidelines. The dairy alter- and flavoured yogurts had very similar sugar contents
natives and drinks categories also had total sugar medians (table 2), fruit yogurts contained small, but appreciably
higher than 5 g/100 g, nonetheless these were still signifi- more, amounts of fibre compared with flavoured or
cantly lower than the children’s, flavoured, fruit and natural/Greek yogurts (0.3 (0, 2.4) vs 0 (0, 0.9) g/100 g).
organic categories with median (range) of 9.2 g/100 g It has previously been reported from an analysis of a US
(0.4–12.5) for dairy alternatives and 9.1 g/100 g (2.3– database that low-fat products, including yogurts, contain
16.5) for drinks (figure 2A). more sugar than their higher fat counterparts.41 However,
Somewhat in contrast to sugar, as seen in figure 2B, many in comparing products below (n=518) and above (n=380)
yogurt categories (children’s, dairy alternative, drinks, the  ≤3 g/100 g threshold for low-fat product labelling,
fruit, natural/Greek) had median levels of fat lower we did not observe this here. Low-fat yogurt products
than the thresholds considered ‘low-fat’ for nutrition had significantly lower amounts of sugar in comparison
claims (3 g/100 g for food and 1.5 g/100 mL for drinks).2 to higher fat yogurts (10.7 (0.1, 21.5) vs 13.1 (0, 32.6)
While flavoured and organic yogurts were just over this g/100 g; figure 3A) and much lower median and tighter

6 Moore JB, et al. BMJ Open 2018;8:e021387. doi:10.1136/bmjopen-2017-021387

 Open access

BMJ Open: first published as 10.1136/bmjopen-2017-021387 on 18 September 2018. Downloaded from http://bmjopen.bmj.com/ on 29 November 2018 by guest. Protected by copyright.
Figure 4  Nutrients in children’s yogurt (n=39) and fromage frais (F Frais; n=62) products. (A) Sugar. (B) Fat. (C) Protein. (D)
Calcium. (E) Energy. (F) Energy/serving. Data were tested for normality and analysed using the Mann-Whitney test. Median is
indicated by black line. **P<0.01; ***P<0.001; ****P<0.0001.

range of fat contents (1.4 (0–3.0) vs 5.5 (1.6–26.7) products marketed to children. Our results highlight that
g/100 g; figure 3B). Low-fat yogurts contained more the total sugar content of yogurts is relatively high in all
protein than higher fat products (4.3 (0, 11.0) vs 3.9 (0.1, categories, with the exception of natural/Greek yogurts.
9.5) g/100 g; figure 3C) and much lower energy per 100 g Very few products, less than 9%, qualified for a low-sugar
(81 (28, 143) vs 125 (36, 445) g/100 g; figure 3D). (≤5 g/100 g) claim for front-of-pack labelling, and almost
While fromage frais is also a fresh lactic fermented none in the children’s category. This is concerning given
milk product, it is made with cheese cultures rather both the continued increase in childhood obesity and
than yogurt cultures. In most of the categories, less than prevalence of tooth decay among children starting school
5% of products were fromage frais and were not sepa- (28%).42 Tooth extractions are shockingly the primary
rated out. However, in the children’s category, fromage reason children aged 5–9 are admitted to hospital with
frais dominated, representing 60% of products (n=62 vs general anaesthetic in the UK.43 Moreover, in 2015/2016
n=39 yogurts) so their nutrient contents were assessed more than one in five children in Reception (ages 4–6),
separately (figure 4). While there was no difference and one in three children in Year 6 (ages 10–12) were
in the sugar content of children’s yogurt and fromage measured as obese or overweight in England.44 While
frais (figure 4A), fromage frais products had lower
yogurt may be less of a concern than soft drinks and fruit
fat (figure 4B; p<0.0001), higher protein (figure 4C;
juices, the chief sources of free sugars in both children
p<0.0001) and higher calcium (figure 4D; p<0.01)
and adult’s diets,39 what is worrisome is that yogurt, as a
contents compared with children’s yogurts. Although
perceived ‘healthy food’, may be an unrecognised source
there was no difference in energy/100 g (figure 4E), the
of free/added sugars in the diet. Indeed, a potentially
energy/serving was significantly lower for fromage frais
(figure 4F; p<0.0001) reflecting its often smaller serving surprising observation from our data was that, after the
size. dessert category, it was organic yogurts that had the
highest median sugar content (13.1 g/100 g). While the
organic label refers to production, the well-documented
Discussion ‘health halo’ effect means that consumers most often
We have comprehensively evaluated yogurt products sold underestimate the caloric content and perceive the nutri-
in the major UK supermarkets in November 2016, exam- tional contents of organic products, including yogurts,
ining nutrient contents across categories, including those more favourably.45

Moore JB, et al. BMJ Open 2018;8:e021387. doi:10.1136/bmjopen-2017-021387 7

Open access

BMJ Open: first published as 10.1136/bmjopen-2017-021387 on 18 September 2018. Downloaded from http://bmjopen.bmj.com/ on 29 November 2018 by guest. Protected by copyright.
An additional challenge for even an educated consumer 20% reduction aim for 11.0 g/100 g product; the report
is understanding that the total sugars on the label include, suggests reformulation, reducing portion size and shifting
in the case of yogurt, the intrinsic milk sugar, lactose, plus portfolio of sales are all viable mechanisms to help achieve
sugars added as sweeteners during processing. UK label- this.37 While median sugar values are not directly compa-
ling laws do not require the declaration of free sugars rable to sales weighted averages, our study suggests the
on nutrition labels and the UK’s sugar reduction guide- organic (13.1 g/100 g), fruit (11.9 g/100 g) and flavoured
lines focus on total sugars for this reason. Interestingly, (12.0 g/100 g) categories require the greatest changes.
although some have argued it is difficult to measure, the Children’s yogurt and fromage frais products had a some-
inclusion of added sugars (under total sugars) on food what lower median of sugar 10.8 g/100 g, possibly meeting
labels has been recently mandated in the USA; companies government guidelines (not clear as not sales weighted).
have until 2020 or 2021 to implement depending on their However, given the recommendations that 4–6 year-olds
size.46 Notably, added sugars in the USA do not include should have no more than 19 g of sugar a day, a single
the sugars in juiced or pureed fruits and vegetables that pot of yogurt can contribute substantially to sugar intakes
are defined as free sugars by the UK and WHO. It has of children. The sugar content of children’s yogurts
been argued for public health purposes, the emphasis and fromage frais varied dramatically and there was no
in communication should be free sugars,47 and it will difference between the sugar contents of yogurts and
be interesting to see how food product companies and fromage frais per 100 g of product. However, as fromage
consumers navigate these changes. Our study highlighted frais has a much smaller serving size (median 47 g vs 90 g
other potential challenges and mixed messages for for yogurt), fromage frais products contained much less
consumers, arising from marketing and packaging. Many sugar per serving than yogurt (5.4 vs 9.2 g sugar/serving);
products that were suggested for children’s lunch boxes with a single serving of yogurt on average delivering close
on supermarket websites were very high-sugar desserts to half of a child’s daily maximal recommended intake of
(from jelly to dairy based) rather than yogurt or fromage sugar.
frais. Retailers could play a positive role in promoting Reformulation is likely to be challenging; beyond acting
health here by establishing boundaries for inclusion in as a sweetener, added sugar in foods acts as a bulking and
lunch box recommendations. Furthermore, the portion colouring agent, and the use of sugar in foods is dictated
sizes for children’s yogurts varied enormously and were by physical and chemical properties that are difficult to
often identical to adult portion sizes. Equally there was substitute.49 In addition, in general consumers ‘liking’
little consistency in portion size in adult yogurts either, for yogurt is correlated positively with sweetness.50 Lactic
and particularly for larger pots (400–500 g) of yogurt the fermentation yields a sour taste that sugar attenuates.
serving size was either not given or was different from the Consumers have been shown to prefer yogurt containing
equivalent smaller pot of yogurts (100–150 g/serving). 10%–13% added sugar but may accept products with
In multiple products with added plant stanols marketed 7% added sugar while rejecting products with 5% or less
for their cholesterol-lowering merits, none would meet added sugars as too sour, or adding sweeteners (caster
a low-sugar claim and several were extremely high in sugar, jam or honey) themselves before consuming.50–53
sugar. In light of data linking high sugar consumption to When Saint-Eve and colleagues53 precisely measured
high cholesterol levels,48 arguably these products should sugar added by 204 French subjects to natural yogurt they
be scrutinised for reformulation with a view to poten- found on average participants added 13.6 g of sugar to
tial added health benefits in terms of obesity and dental their yogurts, more than total content of many commer-
caries. cial sweetened yogurts. Participants underestimated how
The UK guidelines for sugar reduction in the yogurt much sugar they were adding but still perceived their
and fromage frais category do not apply to dairy desserts, addition of sweetener to be the healthier option.53
natural/unsweetened yogurt/fromage frais and yogurt/ Although we have focused our attention on the sugar
dairy drinks (although ‘any sugar-sweetened yogurt content of yogurts per se, it is important to consider
and dairy drinks that are excluded from the soft drinks other nutrients, not least because of the impact on the
industry levy will become part of the sugar reduc- glycaemic response. Notably, our work illustrates that
tion programme’).37 But it was interesting to note the natural/Greek yogurts have a dramatically different
median of sugar content in plain, natural/Greek yogurts macronutrient profile from all other categories,
(5.0 g/100 g), which is predominantly lactose, was higher containing much higher protein (32.4% energy vs range
than the allowance of 3.8 g/100 g that the UK guidelines of 10.8%–20%) and much lower carbohydrate (34.7% vs
agreed on for lactose. It should be noted that the guide- 48.5%–60.4%) than all other categories. The observed
lines have been made, in consultation with industry, on glycaemic index (GI) of yogurts is generally much lower
sales weighted averages so are not directly comparable than predicted values calculated from their carbohydrate
here. This is calculated by weighting the sugar level of contents, with unsweetened yogurts having the lowest
individual products by their volume sales, so that high- GI of all; this is contributed both to the fact that lactose
selling products with high sugar levels will push sales has a low GI, and because yogurts’ protein content will
weighted average upwards. For yogurt, the reported reduce the glycaemic response.54 Similarly, while some
baseline sales weighted average was 12.8 g/100 g, with a studies suggest lactose may be less cariogenic than other

8 Moore JB, et al. BMJ Open 2018;8:e021387. doi:10.1136/bmjopen-2017-021387

 Open access

BMJ Open: first published as 10.1136/bmjopen-2017-021387 on 18 September 2018. Downloaded from http://bmjopen.bmj.com/ on 29 November 2018 by guest. Protected by copyright.
sugars,55 it is recognised that the food components in narrow a time frame as possible and we took only a month
unsweetened dairy products provide a buffering capacity here. However, manufacturers may have made changes to
that is protective to teeth enamel.56 We might underscore, products since this snapshot and it would be interesting
therefore, that it is yogurts sweetened with added sugars to repeat the survey in future years. As nutrient data were
that are primarily of concern here. In particular perhaps, collected from online information, provided by super-
for young children who derive a significant amount of markets or the brand’s own website, these data may not
their free sugar intake, which does not include lactose, always be up to date or accurate. We did not incorporate
from yogurt products (12% for 1.5–4 year-olds and 6% analysis of price here; it would be interesting to assess if
for 4–10 year-olds),39 at a time when taste preferences are sugar or other nutrients relate to price.
being established for life.57 Yogurt without added sugars
remains a nutritious food for all ages and should ideally
be introduced unsweetened during early childhood Conclusions
weaning. Although a drop in daily free sugars intake in While there is good evidence that yogurt can be bene-
the UK has been recently observed, currently consump- ficial to health, products on the market vary widely in
tion is well over 10% in all age groups, much higher than nutrient content. In a comprehensive survey of the UK
the 5% target.39 While clearly in terms of dietary sugar supermarket yogurt products, we highlight here that
intakes and obesity or dental caries, yogurt is not the the median sugar contents of children’s, fruit, flavoured
concern sugar sweetened beverages are, which contribute and organic yogurt categories were well above 10 g/100 g
to 25%–40% of free sugar intake in 1.5–4 year-olds,39 and represented >45% of total energy derived. Organic
nonetheless the rationale for reduction and reformula- yogurts (including organic yogurt with added fruit or
tion across a broad range of products as part of a systemic flavourings, and so on) had the highest median sugar
approach to prevent obesity is robust.34 58 content (13.1 g/100 g). Notably, natural/Greek yogurts
We had preconceived that low-fat yogurts would had a dramatically different macronutrient profile from
contain more sugar than their higher fat alternatives, all other categories, containing much higher protein
in part because of a previous, short report from an (32.4% vs range of 10.8%–20% energy) and much lower
American database analysis that showed low-fat prod- carbohydrate (34.7% vs 48.5%–60.4% energy) contents
ucts, including yogurts, contain more sugar than their than all other categories. While natural/Greek yogurts
higher fat counterparts.41 In contrast, low-fat prod- contained the least amount of sugars, their median
ucts surveyed here had significantly lower sugar than total sugar (5.0  g/100 g, largely lactose) was higher
higher fat products (10.3 vs 13.1 g/100 g in higher fat). than the agreed allowance (3.8  g/100 g) for lactose.
However, we note that although low-fat products did have Although low-fat products had less sugar and energy than
less sugar on average, nonetheless 55% of low-fat prod- higher fat yogurts, nonetheless 55% of low-fat products
ucts had between 10 and 20 g sugar/100 g. With an energy contained 10–20 g sugar/100 g. Within the children’s
intake of 2000 kcal/day, 5% of energy amounts to 25 g category, fromage frais had higher protein and calcium
of sugar, and a single serving of yogurt in all categories contents/100  g than yogurts and was marketed with
surveyed, other than natural/Greek, can easily provide smaller serving sizes. Less than 9%, and only 2% of chil-
half or more of this. On the other hand, a 150 g serving dren’s, products surveyed were low enough in sugar to
of the median sugar (16.4 g/100 g) dessert product could earn ‘green’ in UK front-of-the pack labelling; and many
provide a person the recommended 25 g daily limit of products recommended by supermarkets for lunch boxes
sugar. Nonetheless, low-fat products had lower energy, fat were high-sugar desserts. We conclude not all yogurts are
and sugar contents, and were slightly higher in protein, in as healthy as perhaps consumers perceive them and refor-
comparison to higher fat products. This profile appears mulation for the reduction of free sugars is warranted.
consistent with current UK and US dietary guidelines
that recommend low-fat dairy products out of concerns Contributors  JBM designed the study, analysed the data and wrote the
for obesity and cardiovascular disease. It is notable that manuscript. AH carried out the study, analysed the data and contributed to a
preliminary draft. BAF helped design and interpret the study and revised the
evidence is accumulating that higher fat, rather than manuscript critically for important intellectual content.
low-fat, dairy is associated with a lower risk of obesity and
Funding  This research received no specific grant from any funding agency in the
diabetes,21 23–25 including several studies that suggest full public, commercial or not-for-profit sectors.
fat milk is associated with reduced risk of overweight and
Competing interests  None declared.
obesity in children.25 59 60
Patient consent  Not required.
The study has some limitations. We would have liked to
have included products sold outside the five major online Provenance and peer review  Not commissioned; externally peer reviewed.
supermarkets but this would have created difficulties in Data sharing statement  No additional data are available.
data collections and setting the boundary of inclusion. Open access This is an open access article distributed in accordance with the
However, the supermarkets included covered 75% of the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which
permits others to distribute, remix, adapt, build upon this work non-commercially,
grocery market and is therefore representative of the and license their derivative works on different terms, provided the original work is
overall market, and as such provides a useful database for properly cited, appropriate credit is given, any changes made indicated, and the use
discussion. Ideally data collection should take place in as is non-commercial. See: http://​creativecommons.​org/​licenses/​by-​nc/​4.​0/.

Moore JB, et al. BMJ Open 2018;8:e021387. doi:10.1136/bmjopen-2017-021387 9

 Open access

BMJ Open: first published as 10.1136/bmjopen-2017-021387 on 18 September 2018. Downloaded from http://bmjopen.bmj.com/ on 29 November 2018 by guest. Protected by copyright.
References risk: a systematic review and meta-analysis. Ann Intern Med
1. Thorning TK, Bertram HC, Bonjour JP, et al. Whole dairy matrix or 2014;160:398–406.
single nutrients in assessment of health effects: current evidence and 27. Williams EB, Hooper B, Spiro A, et al. The contribution of yogurt to
knowledge gaps. Am J Clin Nutr 2017;105:1033–45. nutrient intakes across the life course. Nutr Bull 2015;40:9–32.
2. European Parliament Council of the European Union. Regulation 28. Rizzoli R, products D. yogurts, and bone health. Am J Clin Nutr
(EC) No 1924/2006 of the European Parliament and of the Council 2014;99:1256s–62.
of 20 December 2006 on nutrition and health claims made on foods. 29. Salas MM, Nascimento GG, Vargas-Ferreira F, et al. Diet influenced
Official Journal of the European Union 2006:L 404/9. tooth erosion prevalence in children and adolescents: Results of a
3. Morelli L. Yogurt, living cultures, and gut health. Am J Clin Nutr meta-analysis and meta-regression. J Dent 2015;43:865–75.
2014;99:1248S–50. 30. Lythgoe A, Roberts C, Madden AM, et al. Marketing foods to
4. Ashraf R, Shah NP. Immune system stimulation by children: a comparison of nutrient content between children's and
probiotic microorganisms. Crit Rev Food Sci Nutr non-children's products. Public Health Nutr 2013;16:2221–30.
2014;54:938–56. 31. Walker RW, Goran MI. Laboratory determined sugar content and
5. European Food Safety Authority. Scientific Opinion on the composition of commercial infant formulas, baby foods and common
substantiation of health claims related to live yoghurt cultures and grocery items targeted to children. Nutrients 2015;7:5850–67.
improved lactose digestion. EFSA Journal 2010;8:1763. 32. Azaïs-Braesco V, Sluik D, Maillot M, et al. A review of total & added
6. Marette A, Picard-Deland E. Yogurt consumption and impact on sugar intakes and dietary sources in Europe. Nutr J 2017;16.
health: focus on children and cardiometabolic risk. Am J Clin Nutr 33. Moynihan PJ, Kelly SA. Effect on caries of restricting sugars
2014;99:1243S–7. intake: systematic review to inform WHO guidelines. J Dent Res
7. Cormier H, Thifault É, Garneau V, et al. Association between yogurt 2014;93:8–18.
consumption, dietary patterns, and cardio-metabolic risk factors. Eur 34. Moore JB, Fielding BA. Sugar and metabolic health: is there still a
J Nutr 2016;55:577–87. debate? Curr Opin Clin Nutr Metab Care 2016;19:303–9.
8. Zhu Y, Wang H, Hollis JH, et al. The associations between yogurt 35. National Statistics. Statistics on obesity, physical activity and diet,
consumption, diet quality, and metabolic profiles in children in the England, 2017. 2017.
USA. Eur J Nutr 2015;54:543–50. 36. Scientific Advisory Committee on Nutrition. Carbohydrates and
9. Wang H, Livingston KA, Fox CS, et al. Yogurt consumption is health. London: The Stationery Office, 2015.
associated with better diet quality and metabolic profile in American 37. Public Health England. Sugar reduction: achieving the 20%, 2017.
men and women. Nutr Res 2013;33:18–26. 38. Kantar Worldpanel. Grocery market Share for Great Britain 2017.
10. Aune D, Norat T, Romundstad P, et al. Dairy products and the risk https://www.​kantarworldpanel.​com/​en/​grocery-​market-​share/​great-​
of type 2 diabetes: a systematic review and dose-response meta- britain (accessed 24 Nov 2017).
analysis of cohort studies. Am J Clin Nutr 2013;98:1066–83. 39. Roberts C, Steer T, Maplethorpe N, et al. National Diet and Nutrition
11. Gao D, Ning N, Wang C, et al. Dairy products consumption and risk Survey: Results from Years 7 and 8 (combined) of the Rolling
of type 2 diabetes: systematic review and dose-response meta- Programme (2014/2015 to 2015/2016). 2018. https://wwwgovuk/
analysis. PLoS One 2013;8:e73965. government/statistics/ndns-results-from-years-7-and-8-combined
12. Chen M, Sun Q, Giovannucci E, et al. Dairy consumption and risk 40. Department of Health UK. Guide to creating a front of pack (FoP)
of type 2 diabetes: 3 cohorts of US adults and an updated meta- nutrition label for pre-packed products sold through retail outlets.
analysis. BMC Med 2014;12:215. 2016. https://wwwgovuk/government/publications/front-of-pack-
13. Gijsbers L, Ding EL, Malik VS, et al. Consumption of dairy foods and nutrition-labelling-guidance
diabetes incidence: a dose-response meta-analysis of observational 41. Nguyen PK, Lin S, Heidenreich P. A systematic comparison of
studies. Am J Clin Nutr 2016;103:1111–24. sugar content in low-fat vs regular versions of food. Nutr Diabetes
14. Tremblay A, Panahi S. Yogurt consumption as a signature of a 2016;6:e193.
healthy diet and lifestyle. J Nutr 2017;147:1476S–80. 42. Public Health England. National Dental Epidemiology Programme for
15. Wu L, Sun D. Consumption of Yogurt and the incident risk of England: Oral health survey of five-year-old children 2012 A report on
cardiovascular disease: a meta-analysis of nine cohort studies. the prevalence and severity of dental decay, 2012.
Nutrients 2017;9:315. 43. Royal College of Surgeons: Faculty of Dental Surgery. The state of
16. Guo J, Astrup A, Lovegrove JA, et al. Milk and dairy consumption children’s oral health in England, 2015.
and risk of cardiovascular diseases and all-cause mortality: 44. Health and Social Care Information Centre. Statistics on Obesity,
dose-response meta-analysis of prospective cohort studies. Eur J Physical Activity and Diet: England 2017. 2017. https://www.​gov.​uk/​
Epidemiol 2017;32:269–87. government/​statistics/​statistics-​on-​obesity-​physical-​activity-​and-​
17. Sayon-Orea C, Martínez-González MA, Ruiz-Canela M, et al. diet-​england-​2017
Associations between yogurt consumption and weight gain and risk 45. W-cJ L, Shimizu M, Kniffin KM, et al. You taste what you see: Do
of obesity and metabolic syndrome: a systematic review. Adv Nutr organic labels bias taste perceptions? Food Qual Pref 2013;29:33–9.
2017;8:146S–54. 46. Food and Drug Administration. Food labeling: revision of the nutrition
18. Schwingshackl L, Hoffmann G, Schwedhelm C, et al. Consumption and supplement facts labels. Federal Register;81:Docket No.
of dairy products in relation to changes in anthropometric variables FDA–2012–N–1210.
in adult populations: a systematic review and meta-analysis of cohort 47. Mela DJ, Woolner EM. Perspective: total, added, or free? What kind
studies. PLoS One 2016;11:e0157461. of sugars should we be talking about? Adv Nutr 2018;9:63–9.
19. Santiago S, Sayón-Orea C, Babio N, et al. Yogurt consumption and 48. Te Morenga LA, Howatson AJ, Jones RM, et al. Dietary sugars
abdominal obesity reversion in the PREDIMED study. Nutr Metab and cardiometabolic risk: systematic review and meta-analyses of
Cardiovasc Dis 2016;26:468–75. randomized controlled trials of the effects on blood pressure and
20. Guasch-Ferré M, Becerra-Tomás N, Ruiz-Canela M, et al. Total and lipids. Am J Clin Nutr 2014;100:65–79.
subtypes of dietary fat intake and risk of type 2 diabetes mellitus in 49. Chinachoti P. Carbohydrates: functionality in foods. Am J Clin Nutr
the Prevención con Dieta Mediterránea (PREDIMED) study. Am J Clin 1995;61:922S–9.
Nutr 2017;105:723–35. 50. Bayarri S, Carbonell I, Barrios EX, et al. Impact of sensory differences
21. Kratz M, Baars T, Guyenet S. The relationship between high-fat dairy on consumer acceptability of yoghurt and yoghurt-like products. Int
consumption and obesity, cardiovascular, and metabolic disease. Eur Dairy J 2011;21:111–8.
J Nutr 2013;52:1–24. 51. Johansen SB, Næs T, Øyaas J, et al. Acceptance of calorie-reduced
22. Weaver CM. How sound is the science behind the dietary yoghurt: Effects of sensory characteristics and product information.
recommendations for dairy? Am J Clin Nutr 2014;99:1217S–22. Food Qual Prefer 2010;21:13–21.
23. Holmberg S, Thelin A. High dairy fat intake related to less central 52. Chollet M, Gille D, Schmid A, et al. Acceptance of sugar reduction in
obesity: a male cohort study with 12 years' follow-up. Scand J Prim flavored yogurt. J Dairy Sci 2013;96:5501–11.
Health Care 2013;31:89–94. 53. Saint-Eve A, Leclercq H, Berthelo S, et al. How much sugar do
24. Rautiainen S, Wang L, Lee IM, et al. Dairy consumption in consumers add to plain yogurts? Insights from a study examining
association with weight change and risk of becoming overweight or French consumer behavior and self-reported habits. Appetite
obese in middle-aged and older women: a prospective cohort study. 2016;99:277–84.
Am J Clin Nutr 2016;103:979–88. 54. Wolever TM. Yogurt is a low-glycemic index food. J Nutr
25. Beck AL, Heyman M, Chao C, et al. Full fat milk consumption 2017;147:1462S–7.
protects against severe childhood obesity in Latinos. Prev Med Rep 55. Gupta P, Gupta N, Pawar AP, et al. Role of sugar and sugar
2017;8:1–5. substitutes in dental caries: a review. ISRN Dent 2013;2013:1–5.
26. Chowdhury R, Warnakula S, Kunutsor S, et al. Association of 56. Aimutis WR. Lactose cariogenicity with an emphasis on childhood
dietary, circulating, and supplement fatty acids with coronary dental caries. Int Dairy J 2012;22:152–8.

10 Moore JB, et al. BMJ Open 2018;8:e021387. doi:10.1136/bmjopen-2017-021387

 Open access

BMJ Open: first published as 10.1136/bmjopen-2017-021387 on 18 September 2018. Downloaded from http://bmjopen.bmj.com/ on 29 November 2018 by guest. Protected by copyright.
57. Drewnowski A, Mennella JA, Johnson SL, et al. Sweetness and food 60. Bigornia SJ, LaValley MP, Moore LL, et al. Dairy intakes at age 10
preference. J Nutr 2012;142:1142S–8. years do not adversely affect risk of excess adiposity at 13 years. J
58. Dobbs R, Sawers C, Thompson F, et al. Overcoming obesity: an initial Nutr 2014;144:1081–90.
61. U.S. Department of Health and Human Services and U.S.
economic analysis: McKinsey Global Institution, 2014.
Department of Agriculture. 2015 – 2020 dietary guidelines for
59. Scharf RJ, Demmer RT, DeBoer MD. Longitudinal evaluation of milk Americans. 8th edn, 2015.
type consumed and weight status in preschoolers. Arch Dis Child 62. Public Health England. McCance and Widdowson’s the composition
2013;98:335–40. of foods integrated dataset, 2015.

Moore JB, et al. BMJ Open 2018;8:e021387. doi:10.1136/bmjopen-2017-021387 11