• Open Access

Sugar consumption from beverages and the potential effects of a text-based information label


Correspondence to: Jodi P. Gray, Discipline of Public Health, Mail Drop DX 650 550, The University of Adelaide, South Australia 5005; e-mail: jodi.gray@adelaide.edu.au

Although obesity is a multi-factorial and complex problem,1 there is strong evidence that consumption of high-sugar beverages increases the risk of becoming overweight or obese and contributes to the development of diabetes.2–5

Labelling foods with health and nutrition information has been shown to be an effective way to influence consumer perception and to promote behaviour change towards more healthful choices.6–8 We hypothesised that text-based, front-of-pack labels advising individuals of the increased health risks associated with consumption of high sugar beverages would lead to a decrease in purchasing of these beverages.

Informed by a review of relevant literature, the potential impact of two alternative labels was assessed via a postal survey. The full survey documents can be viewed at http://www.adelaide.edu.au/pcsip/drinks. In short, the survey asked respondents about their current purchasing of high sugar and reduced sugar beverages; provided details of a label that would be placed on the front of a range of high sugar beverages; asked a series of questions relating to their comprehension of the label; and then asked respondents to state whether they would alter their current purchasing patterns on the basis of the information on the label.

To target purchasing patterns of households, the postal survey was directed to the person in the household most responsible for grocery shopping. Half the contacted households received a survey stating that the label, ‘Drinking this drink may increase your risk of obesity’ was to be placed on sugar-sweetened soft drinks, fruit juices, fruit drinks and cordials. The remaining contacted households received an alternative label, ‘Drinking sugar sweetened drinks may increase your risk of diabetes’, to be placed on sugar-sweetened soft drinks, fruit drinks and cordials (but not fruit juices). Neither label was to be placed on flavoured milk.

Of 197 surveys delivered, 130 (66%) were returned, 68 (52%) contained the obesity label. Table 1 describes current consumption patterns. 79.2% of households purchased at least one of the five high sugar beverages. In households that purchased high sugar beverages, mean (median) sugar intake per person per week from these beverages alone was 314 g (185 g), assuming equal consumption across the household. This is equivalent to an energy intake of 762 kJ (450 kJ) per person per day. The World Health Organization recommends that free sugars make up no more than 10% of the total energy intake for adults.9 Based on recommended total energy intake (8700 kJ),10 21.3% (95% CI 14.0–28.6) of all households exceeded this level from high sugar beverage consumption alone.

Table 1.  Current beverage purchasing in a typical week (per person in household).
 Number PurchasingVolume purchased (L)a Sugar Intake (g)a  
Beverage Type(n=130)(SD)   (g/100 mL)(SD)   
  1. a) Includes only households which purchase the beverage. Excludes households missing volume or household data. Per person quantities calculated as household volume divided by total number of people in household.

Regular soft drink52 (40.0)2.2 (2.6) (280.4)143.76.71,293.6
Fruit drink17 (13.1)1.0 (0.7) (74.2)105.819.8264.5
Regular cordial36 (27.7)0.7 (0.5) (210.8)233.912.5802.0
      (8.0 diluted)    
100% fruit juice78 (60.0)0.9 (0.5) (52.0)69.412.3261.3
Flavoured Milk22 (16.9)1.3 (2.1) (203.6)47.817.9764.8
All high sugar beverages103 (79.2)2.4 (2.8) (371.1)185.112.32,307.2

Table 2 shows that of the 100 households who purchased one or more high sugar beverages, 36.0% (CI 26.6–46.2) stated that they would reduce purchasing of one or more high sugar beverages. Of those receiving the obesity label, 35.3% stated they would reduce purchasing of at least one of the four labelled high sugar beverages (CI 22.4–49.9). For the diabetes label, 41.7% (95% CI 25.5–59.2) stated they would reduce purchasing of at least one of the three labelled high sugar beverages. 13% of households (CI 7.1 to 21.2) would reduce purchasing of all currently purchased high sugar beverages.

Table 2.  Stated change in purchasing for high sugar beverages currently purchased.a
 % who would reduce volume purchased (95% CI)
Beverage TypeBothObesityDiabetes
  1. a) n=100. 3 households did not respond to change in consumption question.

  2. b) Obesity label in obesity version of survey.

  3. c) Diabetes label in diabetes version of survey.

  4. d) No label.

  5. e) One-sided, 97.5% confidence interval.

Regular soft drinkbc37.5 (24.0–52.6)31.6 (12.6–56.6)41.4 (23.5–61.1)
Fruit drinkbc30.8 (9.1–61.4)44.4 (13.7–78.8)0.0 (0.0–60.2) e
Regular cordialbc27.3 (13.3–45.5)18.8 (4.0–45.6)35.3 (14.2–61.7)
100% fruit juiceb18.3 (10.1–29.3)22.7 (11.5–37.8)11.1 (2.4–29.2)
Flavoured milkd16.7 (3.6–41.4)0.0 (0.0–28.5) e42.9 (9.9–81.6)
At least 1 of the 5 beverages36.0 (26.6–46.2)34.6 (22.0–49.1)37.5 (24.0–52.6)

It is interesting to note that households receiving the diabetes label appear to have extrapolated the warning from the labelled high sugar beverages to flavoured milk, but not fruit juices. A similar effect on purchasing of flavoured milk was not noted for the obesity label.

The current study provides preliminary evidence that a front-of-pack text-based information label advising consumers of the health risks associated with high sugar beverages is a potentially effective way to reduce purchasing of these beverages. The main limitation of the current study is that participants report intended reductions in the purchase of high sugar beverages, reductions are not observed. Supporting evidence is derived from a study that assessed the effectiveness of text-based information labels on Philadelphia cream cheese, which led participants to choose to taste lower fat versions.8

The Food Labelling Law and Policy Review,11 initiated by COAG, is currently reviewing the evidence on food labelling in Australia and New Zealand. The Review has sought consultation on whether warnings should be placed on food labels indicating health risk (e.g. high saturated fat per serve) and the related health consequences. Available evidence on the effectiveness of text-based information labels suggests that they should be considered as alternatives to other front-of-pack labelling systems (e.g. % Daily Intake, UK Traffic Light System).