A new method for estimating added sugars
Including added sugars on food labels poses a technical challenge because natural and added sugars are chemically identical and therefore cannot be reliably or practically measured outside a laboratory. There have been a number of methods proposed to estimate added sugars, starting with the 10-step method developed by Louie et al at Sydney University we covered in Sweet Bites back April 2016. The Australian Bureau of Statistics (ABS) used this method to develop a new food database, and their Australian Health Survey added sugars consumption report (2011-12).
Now a group from Cambridge University (Amoutzopoulos et al 2018) have proposed a new recipe disaggregation methodology with only 5 steps to estimate added and free sugars from the UK National Diet and Nutrition Survey (UKNDNS). We asked some of our own local experts for their view on it.
STEP 1: Ingredients and their proportion are identified from the ingredients list and nutrition information on the label.
STEP 2: A total sugar value is added to each ingredient per 100g, obtained from composition tables/databases.
STEP 3: Total sugars from each ingredient is assigned to one of the sugar components.
STEP 4: The proportion of each ingredient is used to calculate a value in grams for each sugar component using a formula.
STEP 5: The grams of sugar components in 100g of the food are then added to give estimated added sugars and free sugars per 100g.
The method can be adjusted to estimate International definitions of sugars.
Added sugars (EFSA) = sugar-table + sugar-other
The European Food Safety Authority (EFSA) defines added sugars as sucrose, fructose, glucose, starch hydrolysates (glucose syrup, high-fructose syrup) and other isolated sugar preparations added during food preparation and manufacturing and does not capture the sugars present in unsweetened fruit juice or honey.
Free sugars (WHO) = sugar-table + sugar-other + sugar-honey + sugar-fruit juice
WHO free sugars refer to all sugars that are added during food manufacturing and preparation as well as sugars that are naturally present in honey, syrups, fruit juices, and fruit concentrates.
Free sugars (SACN) = sugar-table + sugar-other + sugar-honey + sugar-fruit juice + sugar-pureed fruit
Scientific Advisory Committee on Nutrition (SACN). As a variant to the WHO definition, the UK has adapted the ‘free sugars’ definition to exclude the ‘lactose’ in milk and the sugars within the cellular structure of foods (particularly fruits and vegetables) - this distinction is not specified by WHO.
NMES (non-milk extrinsic sugars) = sugar-table + sugar-other + sugar-honey + sugar-fruit juice + 50% sugar-stewed fruit + 50% sugar-dried fruit
NMES is a term used for some time in the UK and includes the sugars not contained within the cellular walls of plants, all sugars added to foods, and half the sugars in canned, stewed, dried or preserved fruits.
How accurate is it?
To prove the rigour of their method, they performed an inter-rater repeatability test to see if two different researchers would arrive at the same sugar values for a set of 50 foods. There were overall no statistically significant differences, although there were some differences between sugar types. The largest differences were in ‘sugar-other’ values and this was expected as there is no way to distinguish other sugars from table sugar from food label information. Thirty-five foods out of 50 had less than 1g difference between added and free sugar values. The main differences (albeit small) arose from understanding the product properties such as identifying sugar-rich ingredients, allocating a sugar type, use of moisture loss factors and identifying the food processing as it pertained to sugar type, such as boiled, roasted and stewed fruit. Foods with multiple sugar-rich ingredients: breakfast cereals with fruit, fruit snacks, posed the biggest challenge.
The authors conclude their method allows food specific estimations for a variety of sugar definitions and is efficient and transparent. They add it could allow food composition databases to add and update sugar values, which would then help measure population intakes from dietary surveys.
Are there any down sides?
The method is complicated and requires technical skill and expertise to calculate. It also relies on knowing the ingredients and their proportions in every food. In the case of home-prepared or café/restaurant foods, it relies on having a standardised recipe when, recipes can vary. In the case of commercially produced packaged foods the method relies on knowing the exact formulation - not available on the label - and this would need to be provided by the manufacturer. This presents a challenge for small manufacturers without skilled staff, and a challenge for maintaining currency because formulations change.
What do local experts think?
Notwithstanding the fact the Sydney method estimates added sugars and the Cambridge method estimates several definitions of sugar, we asked a few experts what they thought, and how the methods compared for accuracy and utility.
Dr Jimmy Louie, the first author of the original Sydney University group paper proposing a method for estimating added sugars thinks the Cambridge method is more useful for academic research than any practical purpose, such as food labelling. He says,
“The Cambridge method requires much more information than my 10-step method and therefore the accuracy should increase as a result. However, it’s a challenge to obtain the proprietary recipes of packaged foods and this limits its practical use”.
Professor Jennie Brand Miller from the University of Sydney had a more fundamental question of why we would want to estimate free sugars in the first place. She says estimating free sugars “doesn’t make too much sense because it’s not proven juices do harm. It’s not science”.