Previous Issues

How accurate are Nutrition Information Panels when it comes to Nutrient Variability?

Published: 15 Mar 2007

by Joe Lederman, John Gao and Tony Zipper © Lawmedia Pty Ltd, March 2007
Australian Food Lawyers & Consultants

Nutrient quality can vary significantly, even between different items of the same food.  What usefulness does a Nutrition Information Panel have if the law permits the inclusion of data that ignores huge variability in nutrient profile data? Is the food law that requires a Nutrition Information Panel really meaningful to inform consumers about what is nutritious or healthier food? How scientific is the approach of the law, or are Nutrition Information Panels merely intended to blind consumers by science?

The Australia New Zealand Food Standards Code (“ANZFSC”) requires in Standard 1.2.8 that packaged foods be sold with a Nutrition Information Panel on the label.

The User guide issued by the government food agency, Food Standards Australia and New Zealand (FSANZ), provides interpretative guidelines to Standard 1.2.8 on Nutrition Information requirements.  The FSANZ User guide explains that there are four alternative methods to obtain the information required to produce a Nutrition Information Panel. These alternative methods are:

  1. Nutrition Panel Calculator (a tool provided by FSANZ on its web site)
  2. Laboratory Analysis
  3. Food composition tables or databases
  4. Other software interfaces for food composition tables and databases.

Other than through actual laboratory analysis mentioned above, which involves actual analysis of the physical and chemical properties to determine all the required nutrients (other than the figures for total carbohydrates, which are to be calculated in accordance with one of 2 prescribed methodologies under Clause 1 of Standard 1.2.8), the other three methods are all based on empirical data accumulated from various sources, all of which have been sourced originally from some laboratory analysis.

Nevertheless, all such data irrespective of its source would be subject to being an average or mean of data accumulated over many years from many analyses. For some inexplicable reason, the law always assumes that this data has been derived from the same named biological source, even though this assumption is a fiction.

This raises the question as to how this accumulated data can relate to the same produce that was used to manufacture a particular food today and again in a week’s time for in three months time or for any additional batch at the end of the shelf life span of the previous batched product.

Surely, all natural produce is subject to variations due to soil conditions, amount of water, weather conditions and treatments applied by the farmer and subsequent handling in the supply and storage chain. These variations would obviously have an effect on yield, size of the ripe produce, and the nutrient contents of that grown food.  Having regard to the current debate on “climate change” (i.e. drought, elevated temperatures, etc), would this not affect nutrient profiles too? How will these variations in the nutrient profiles of farm produce today be reflected in the calculation of the values shown in Nutrition Information Panels of the processed products at a later date?

Each of the following three introductory quotations are cited from the FSANZ document “Explanatory Notes for the Use of the Nutrition Panel Calculator (NPC) – Supported by AUSNUT Special Edition (3) (Australian Food and Nutrient Database for Nutrition Labelling – Release 3) October 2004”:

“7.1 Limitations of the data:

“There are a number of limitations inherent with all food composition data, including those presented in the NPC. Food composition data are mostly estimates that attempt to provide representative data. Foods, being biological materials, vary greatly in their natural nutrient composition. For primary produce, the variability in nutrient composition may be due to different methods of plant and animal husbandry, storage, transport and marketing.”

“It is important to note that the data presented on the NPC were not developed for the purpose of calculating nutrition information panels.”

“You should be aware that given these limitations, the results generated by the NPC will only ever be approximate in nature, rather than an exact reflection of your product’s average nutrient composition.”

The above three quotations constitute disclaimers by FSANZ. They amount to confessions that challenge the factual and legal basis for consumers to be relying on the figures produced in every Nutrition Information Panel.  If so, the law opens itself to an argument that inaccuracy in the Nutrition Information Panel can be acceptable.

It should be noted that the first of these three disclaimers is also similarly expressed in the introduction to other popular and generally accepted data sources, such as:

Nutritional Values of Australian Foods by Ruth English and Janine Lewis 1991.

The Concise New Zealand Food Composition Tables (4th Edition) by Athar, Spriggs & Lui, 1999.

U.S. Department of Agriculture, Agricultural Research Service. 2006. USDA Nutrient Database for Standard Reference, Release 19. Nutrient Data Laboratory

McCance and Widdowson’s The Composition of Foods – 6th Edition as jointly published by The Royal Society of Chemistry and the Ministry of Agriculture, Fisheries and Food.

If these variability factors are to be expected according to the government authorities that are responsible for the setting of food standards and the enforcement of food laws, then should the manufacturer be held responsible for substantial variability in nutrient values data?  On the other hand, what use is legally sanctioned inaccuracy of Nutrition Information Panels to consumers?

Since it is mandatory by law to require that the Nutrition Information Panel accompany all food products on the label of packaged food or be available on request for the buyer of unpackaged food, what is the point of having the law exist at all?

The relevant State and Territory health authorities and local government offices are responsible to ensure that the nutrition information provided is accurate and not deceptive to consumers. Yet how enforceable can the law be expected to be if the laws governing Nutrition Information Panels recognize that the nutrient and nutrition information can be inherently inaccurate and yet be legally acceptable?

In 2004, two reports appeared that discussed the accuracy of food labeling in relationship to the Australia New Zealand Food Standards Code (ANZFSC):

  1. FSANZ Food Label Monitoring Survey – Phase 1 Report: This report covered a survey on all labelling aspects including the use and format of Nutrition Information Panels. However, this report did not address the accuracy of the actual nutrients expressed in the Nutrition Information Panels.
  2. Western Australian Food Monitoring Program – Accuracy of Nutrition Information Panels 2000-2002. The Executive Summary of this report stated:
    “This survey was conducted by the WAFMP to examine products specifically making claims of low (reduced) fat, low salt, high carbohydrates or no claim – but still carried a Nutrition Information Panel to compare results against. The results have been broadly grouped according to the claim on the label.  The majority of the results were within 25% of label declarations across all Nutrition Information Panel characteristics, however the level of inaccuracies are significant, particularly in products claiming ‘low fat’ and are of relevance to enforcement agencies who undertake compliance monitoring.”

(It should be noted that this survey was of products during 2000- 2002, i.e. prior to the introduction of mandatory NIP labelling of all packaged food).

This Western Australian report offered many details of the accuracy levels of many nutrients, i.e. label claim versus analysis. The following quotation is from page 12 of the said WA report:

4.3 Results for products making no nutrient claims

Results from this grouping of products revealed Energy and Total carbohydrates to be the most accurate nutrient declarations.

Protein levels were found to be reasonably accurate with 73% of products being within 25% of the declared values.

Sodium and fat were fairly inaccurate, with only 63.4% and 53.6% (respectively) of samples tested within 25% of declared values.

Sugars were the least accurately declared nutrient, with only 50% of samples found to contain the nutrient at levels greater than 75% variance from the declared values.”

As it not known whether the declared values in the NIP were determined by analysis of representative samples or calculated from data, it is not possible to draw a conclusion as to how such variations actually occurred.

It is of interest that the Australian Food Standards Code originally (1987) permitted (in what is now referred to as Volume 1 -Australia Food Standards Code, in clause A1.13 (h)) that the value shown in a Nutrition Information Panel was deemed to comply if these values (of energy, carbohydrate, starch or fibre) did not vary by more than 20% from those values actually present and 10% variation was permitted for other nutrients. This permission was withdrawn by Amendment 11 in August 1991. However, even though these figures are no longer part of the law, they are still regarded as part of the folklore of the food industry and seem to be still regarded as the de facto permitted limits.

The current Australia New Zealand Food Standards Code does not permit or mandate any limits on accuracy of the levels of nutrients expressed in Nutrition Information Panels but only requires that these values be ‘average’ values. Maximum and minimum quantities are required in regard to claims for polyunsaturated and monounsaturated fatty acid contents of food.

The extent and the legal impact of inaccuracy of information shown on Nutrition Information Panels have not, to the writers’ knowledge, been researched or reported widely. 

The issue of inaccurate nutrition information on food labels did surface in late November 2006 following a media release by the NSW Food Authority and attracted publicity in the media.

To the date of this article, we believe the Australian Competition and Consumer Commission (ACCC) has not tested or challenged the inaccuracy of NIPs although there have been cases where action has been taken by the ACCC for incorrect statements of nutrients but not specifically in connection with the offence of having an inaccurate Nutrition Information Panel.  An example is the Nestle case in 1997 where Nestle was found to have incorrectly stated the sugar levels in “Vitari”.  Similarly, the ACCC has prosecuted cases against manufacturers who have diluted fruit juices with extra water.  Although implicitly, the Nutrition Information Panels of these products would have been inaccurate, the inaccuracy of the figures within the NIP per se was not the basis for prosecution. Author’s note: Since this article was written, the ACCC has public expressed concern over the accuracy of nutrition information panels on food products. On 21 March 2007, the ACCC expressed concerns over the accuracy of a Vitamin C content claim made on nutrition labels attached to Ribena blackcurrant fruit drink products (manufactured by GlaxoSmithKline Australia Pty. Ltd.), emphasising the need for companies to use appropriate calculation methods when making nutrition content claims. GlaxoSmithKline has provided court enforceable undertakings to redress this issue but is still facing a New Zealand prosecution case over the same issue in New Zealand. The ACCC media release can be viewed here