ANNEXES

ANNEX I: PARTICIPANTS - TECHNICAL WORKSHOP ON FOOD ENERGY: METHODS OF ANALYSIS AND CONVERSION FACTORS

Principal participants

Junshi Chen
Institute of Nutrition and Food Safety
Chinese Center for Disease Control and Prevention
29 Nan Wei Road
Beijing 100050, China

Simon Chevassus-Agnès
Les Molunes
F-39310 France

G. Sarwar Gilani
Senior Research Scientist
Nutrition Research Division, Food Directorate
Health Products and Food Branch
2203C Banting Research Centre
Ross Avenue
Ottawa,Ontario K1A 0L2, Canada

James Harnly
USDA, ARS, BHNRC, FCL
Bldg. 161, BARC-East
Beltsville, MD 20817, USA

Geoffrey Livesey
Independent Nutrition Logic
21 Bellrope Lane
Wymondham, Norfolk NR18 0QX, United Kingdom

William C. MacLean , Jr. (Chairperson)
Vice President, Medical and Regulatory Affairs
Ross Products Division
Abbott Laboratories
Columbus, Ohio 43215, USA

Basil Mathioudakis
Principal Administrator
Food Law and Biotechnology
European Commission
Health and Consumer Protection Directorate-General
Rue de al Loi/Wetstraat 200
B-1049 Brussels, Belgium

Miriam Muñoz de Chavez
Centro de Investigación de Ingenieria y
Ciencias Aplicadas (CIICAP)
Facultad de Medicina
Universidad Autónoma del Estado de Morelos
Av. Universidad, 1001
Col. Chamilpa
62210 Cuernavaca, Morelos, Mexico

Benjamin Torun
CIDAL
15 Ave 14-79, Zona 13
Guatemala City, Guatemala

Mauricio T.L. de Vasconcellos
Fundação Instituto Brasileiro de Geografia e Estatística (IBGE)
Departamento de Metodologia
Av. República do Chile, 500/10° andar
20031-170 Rio de Janeiro, Brazil

Penelope Warwick (Rapporteur)
School of Biological, Biomedical and Molecular Sciences
University of New England
Armidale N.S.W. 2351, Australia

Other participants

François Sizaret
21 Via Montagne Rocciose
00144 Roma, Italia

Gina Kennedy
Consultant- ESNA, FAO
Via San Giovanni in Laterano 22
00184 Rome, Italy

World Health Organization

Zita Weise Prinzo
Technical Officer
Department of Nutrition for Health and Development
World Health Organization
20, Avenue Appia
CH-1211 Geneva, Switzerland

SECRETARIAT

FAO Economic and Social Department, Food and Nutrition Division (ESN)

Kraisid Tontisirin
Director
Food and Nutrition Division

Prakash Shetty
Chief
Nutrition Planning, Assessment and
Evaluation Service (ESNA)

Barbara Burlingame
Senior Officer
ESNA

Ruth Charrondière
Nutrition Officer
ESNA

Marie-Claude Dop
Nutrition Officer
ESNA

Robert C. Weisell
Nutrition Officer
ESNA

FAO Economics and Social Department, Statistics Division (ESS)

Jorge Mernies
Chief
Statistical Analysis Service
ESSA

Ricardo Sibrian
Statistician
Statistical Analysis Service
ESSA

Codex Alimentarius Commission Secretariat (ESNC)

Selma Doyran

Food Standards Officer

Secretariat, Codex Alimintarius Commission, ESNC

ANNEX II: MEMBERS OF WORKING GROUP 5, THEIR RECOMMENDATIONS AND THE MODIFICATIONS TO THOSE RECOMMENDATIONS MADE BY THE CURRENT TECHNICAL WORKSHOP PARTICIPANTS

As part of the Expert Consultation on Protein in Human Nutrition that took place in July 2001, a working group - Working Group 5 - met for two days to address Analytical Issues in Food Energy and Composition: Energy in Food Labelling, including Regulatory and Trade Issues. This annex gives their recommendations, which were extensively considered and modified to some degree by the current technical workshop participants. These modifications are also given.

1. Members of Working Group 5

Chairperson

* Ghulam Sarwar, Research Scientist, Health Protection Branch, Health Canada, Ottawa, Ontario, Canada

Members

Malcolm Fuller, 107 Quaker Path, Stony Brook, New York, United States

* Geoff Livesey, Independent Nutrition Logic, Wymondham, Norfolk, United Kingdom

Paul Moughan, Professor, Institute of Food, Nutrition and Human Health, Massey University, Palmerstown North, New Zealand

Peter Pellet, Professor Emeritus, University of Massachusetts, Department of Nutrition, Amherst, Massachusetts, United States

Paul Pencharz, Senior Scientist and Professor, The Hospital for Sick Children/University of Toronto, Toronto, Canada

Secretariat

* Barbara Burlingame, Senior Officer, Nutrition Planning, Assessment and Evaluation Service (ESNA), Food and Nutrition Division, FAO, Rome, Italy

J.H. Jones, Professor, School of Dietetics and Human Nutrition, McGill University, Montreal, Canada

Daniel Tomé, Professor in Human Nutrition, Institut National Agronomique Paris-Grignon (INA, P-G), GER Biologie et Nutrition Humaines, Paris, France

* Also participated at the current technical workshop.

2. Recommendations of Working Group 5 - Analytical Issues in Food Energy and Composition: Energy in Food Labelling, including Regulatory and Trade Issues

Summary

The Working Group met to review and consider the literature on this topic, with special reference to: a) the routes of energy loss from the body such that they could not normally be used to maintain energy balance; b) the size of each energy loss for the energy-providing substrates, including fermentable carbohydrates; c) variation in energy losses among studies of food components; d) the fact that traditional foods have associated energy losses that are not taken into account at present; and e) the factors that modulate energy requirements. Various possible approaches to energy evaluation were examined against a set of criteria suggested in the working paper for the selection of energy values. These included approaches that would account for substrate-associated thermogenesis. The Working Group was aware of difficulties that could arise in the use of various terms that have been used to describe food components to which food energy factors are applicable, and sought to make clarifications. The following recommendations were made.

Energy conversion factors

Recommendation 1

Consideration should be given to the use of the net metabolizable energy (NME) system of factors for food labelling, for food tables and when calculating practical food needs from energy requirements.

Recommendation 2

Circumstances external to food energy availability should be considered to vary energy requirements and not to vary energy availability. Environmental temperature, drugs, exogenous hormones and bioactive compounds should be considered to act on the utilization of NME.

Recommendation 3

The SI unit of energy is the Joule, which should be used for the expression of energy in nutrition (J, kJ, MJ, etc.), without needing to be accompanied by the calorie equivalent.

Terminology

Recommendation 4

To avoid confusion with the National Research Council’s definition (NRC, 1981) of net energy, the contraction of “net metabolizable energy” to “net energy” is not advised.

Recommendation 5

INFOODS Tagnames (Klensin et al., 1989) should be used for the unambiguous identification of food components in food databases and for other purposes as appropriate.

Recommendation 6

• Available carbohydrate is a useful concept in energy evaluation and requires to be retained.

• For energy evaluation purposes, an assured direct analysis of available carbohydrate is considered preferable to an assessment of total carbohydrate by difference less dietary fibre.

• Nevertheless, assessment of total carbohydrate by difference less dietary fibre is considered acceptable for energy evaluation purposes with traditional foods, but not with clinical formulations or novel foods, or with foods when reduced or low-energy claims are made.

• In energy evaluation, it is preferable to express available carbohydrate as monosaccharide equivalents.

• The term “glycaemic carbohydrate” may be confused with glycaemic index and has no use in energy evaluation. The term “glycaemic index” should inform the consumer about the quality, and not the quantity, of carbohydrate in foods.

Recommendation 7

• The term “dietary fibre” is familiar to consumers and ought to be retained in food labelling, and therefore in food tables.

• The concept of soluble versus insoluble dietary fibre should not be used in energy evaluation.

• For the present, the AOAC (2000) - Prosky (985.29) method of dietary fibre analysis or similar method should be used in food analysis for the purpose of calculating food energy.

• Fermentability of dietary fibre should be adopted in energy evaluation.

• In determining the energy factor for dietary fibre in traditional foods, a general fermentability value of 70 percent has been assumed when deriving the energy factors of 8 kJ ME/g and 6 kJ NME/g. However, when a nutritional claim for energy is made, the fermentability of the dietary fibre should be determined.

• When unavailable carbohydrate in a product or ingredient is quantifiable as non-starch polysaccharide, fermentability determinations can use Englyst (AOAC, 2000) or similar methodology.

• Isolates of dietary fibre or foods that are artificially enriched in any components of dietary fibre (e.g. resistant starch) should have their specific energy factors determined.

• At present, dietary fibre terminology confuses a number of analytical methods. The results of each method should appear in separate columns in food tables, identified by INFOODS tagnames, allowing energy factors to be applied as appropriate.

Recommendation 8

• For energy evaluation purposes it is recommended that fats are analysed as fatty acids and expressed as triglycerides.

• Poorly or non-available fats and fat-like substances (e.g. artificial fats and natural waxes) and fats that are unusually high in selected fatty acids (e.g. when compared with traditional foods) should have specific energy values if the values differ significantly from the general energy value of fats.

• The non-digestible fat content of poorly or non-digestible fats should not contribute to the nutrient fat content of a food.

Recommendation 9

The term “sugar alcohol” should be phased out of food labelling and replaced with “polyol”. Polyols should be recognized as carbohydrates, but not sugars.

Recommendation 10

Individual energy factors for the most abundant organic acids should be required in energy evaluation, although a general factor can usually be applied in food labelling.

3. The current technical workshop’s modifications to the recommendations of Working Group 5

Among bullets in Recommendation 6

• For energy evaluation purposes, a standardized direct analysis of available carbohydrate is considered preferable to an assessment of total carbohydrate by difference less dietary fibre.

Among bullets in Recommendation 7

• In determining the energy factor for dietary fibre in traditional foods, a general fermentability value of 70 percent was adopted (Livesey, 1990) when deriving the energy factors of 8 kJ ME/g and 6 kJ NME/g. Specific energy factors should apply to added and novel (EC, 1997) or functional (FNBIM, 2002) fibre (e.g. resistant starches, inulin, polydextrose).

• At present, dietary fibre is determined by a number of methods yielding different results. Therefore, the results of each method should appear in separate columns of food tables, identified by INFOODS tagnames, allowing energy factors to be applied appropriately.

Among bullets in Recommendation 8

• Determination of total lipid by standardized gravimetric methods is acceptable for most energy evaluation purposes. However, it is preferred that fats are analysed as fatty acids and expressed as triglycerides because this excludes waxes and the phosphate content of phospholipids.

• The non-digestible triglyceride or fat content of energy-reduced fats should not contribute to the nutrient fat content of a food.

Note: Modifications were also suggested regarding labelling of the nutrient content of carbohydrate.

ANNEX III: CORRECTIONS TO THE DIET AND/OR STANDARD ENERGY REQUIREMENTS WHEN USING METABOLIZABLE ENERGY (ME) OR NET METABOLIZABLE ENERGY (NME) FACTORS

Energy requirement estimates for all ages are currently based on representative energy expenditures plus additional needs for growth, pregnancy and lactation. Once the energy requirement has been estimated, the food energy intakes needed to match this requirement must be determined. Current estimates of energy requirements are based on methods that reflect the measurement of metabolizable energy (ME) expenditure and hence are expressed in ME terms (FAO, 2004). However, energy expenditure varies with the composition of the diet, particularly with protein, alcohol, fibre and other fermentable compounds and short- to medium-chain fats. Because current estimates of requirements were derived from healthy people consuming representative diets, a correction may be desirable in circumstances in which the composition of the diet departs from “average”. In addition, there are some clinical situations in which such a correction may prove useful.

Methods of applying corrections when matching intakes with requirements were previously given for variations in dietary fibre content (WHO, 1985). A revised method to correct for the composition of the diet using ME is shown in Box III.1, while an alternative method using net metabolizable energy (NME) is shown in Box III.2. The basis for the corrections is outlined in the current report. For some purposes, corrections may not be necessary. For regular diets, i.e. those consumed by 95 out of every 100 people in an adult group - male, female, of any age, healthy, non-slimming - and with no more than average alcohol consumption, the correction will usually be less than 2.5 percent (Livesey, 2002). For practical purposes, diets containing 10 to 20 percent of energy (ME) from protein, 1 to 3 percent of energy (ME) from fibre and 0 to 6 percent of energy (ME) from alcohol result in errors of less than 2.5 percent.

An alternative approach uses revised food energy factors (NME values) and a single adjustment of energy requirement to the same scale. This method overcomes the problem of regular diets with similar ME contents differing from one another by up to 5 percent in their capacity to match energy requirements, and by up to 11 percent in the protein intakes that fall within the range from adequate to tolerable upper intakes (Livesey, 2002). The method also overcomes the problem of some traditional foods and novel ingredients with similar ME contents differing by up to 30 percent in their capacity to contribute to the standard energy requirement estimate (Livesey, 2001).

ANNEX IV: COMPARISONS OF ENERGY CONTENTS OF BREASTMILK, INFANT FORMULA AND SELECTED FOODS FOR INFANTS AND YOUNG CHILDREN USING ME AND NME ENERGY CONVERSION FACTORS^[17]

The energy content of breastmilk is calculated in four different ways in Table 3.6 (see Section 3.8 on p 37) using representative values for protein (8.9 g/litre), fat (32 g/litre) and carbohydrate (74 g/litre) (Fomon, 1993). ME-ATW represents an estimation of the energy content using standard Atwater general factors. ME-specific is calculated using the Atwater specific factors published by Merrill and Watt (1973). NME-1 and NME-2 values are calculated using NME factors with two different sets of assumptions. The NME-1 value was calculated using the same figures for protein, fat and carbohydrate as were used to calculate ME-ATW, and does not take into account the different digestibilities of immunoglobulins and oligosaccharides in human milk. Immunoglobulins comprise about 12 percent of the protein in human milk - about 1.1 g/liter (Fomon, 1993) - mostly (95 percent) IgA which, along with some other proteins in human milk such as lactoferrin, is not completely digestible. Up to 10 percent of the protein in human milk may not be nutritionally available to the infant (Davidson and Lonnerdal, 1987). Oligosaccharides comprise about 15 percent of carbohydrate in breastmilk (13 g/litre) (McVeagh and Miller, 1997; Coppa et al., 1997). These carbohydrates are not digestible, but are fermentable in the colon. NME-2 in Table IV.1 is a recalculation of the energy content of human milk assuming that 10 percent of the protein is unavailable and applying the NME factor for unavailable carbohydrate to oligosaccharides. The energy value calculated with the Atwater specific factor value is about 3 percent lower than that calculated using Atwater general factors. NME results in a decrease of an additional 4 percent, regardless of the assumptions made.

Table IV.1 examines the effects of using NME conversion factors rather than Atwater general or specific factors on the energy contents of a standard milk-based formula^[18] containing 14 g/litre protein (non-fat milk and whey protein concentrate), 36.5 g/litre fat (a mixture of three vegetable oils) and 73 g/litre carbohydrate (lactose). Heat processing of lactose-containing formulas converts some of the lactose to lactulose (Beach and Menzies, 1986; Hendrickse, Wooldridge and Russell, 1977). Lactulose is indigestible in the small intestine, but fermentable in the colon. Levels greater than 3 g/litre are uncommon. NME-2 shows the recalculated energy content of the formula assuming a level of lactulose of 3 g/litre and a corresponding decrease of lactose content. Using NME rather than ME (Atwater general factors) results in a decrease of between 5 and 6 percent.

A representative soy protein-based infant formula^[19] is also shown in Table IV.1. Protein (16.6 g/litre) is provided by soy protein isolate and l-methionine, fat (36.9 g/litre) by a mixture of vegetable oils, and carbohydrate (69.6 g/litre) by a mixture of corn syrup solids and sucrose. The energy contents using Atwater general and specific factors and NME factors are shown. Soy protein-based formulas have an inherent content of fibre - soy polysaccharide - that derives from the soy protein isolate ingredient used in their manufacture. This fibre, which would not normally exceed 3 g/litre, is fermentable to some degree. NME-3 in the table assumes this level of soy polysaccharide and applies the NME value for unavailable carbohydrate. Use of NME factors results in a decrease of between 4 and 5 percent compared with ME general factors.

Tables IV.2A and IV.2B show the composition and declared values for energy for four representative baby foods from a single manufacturer^[20] and compare these values with those calculated using Atwater general and, where possible, specific factors and NME factors. Using NME rather than ME general factors results in variable decreases in energy content from as low as 2 percent for rice cereal to as high as 9 percent for chicken with gravy.

TABLE IV.1
Energy values of standard infant formulas

¹ Similac(with Iron (Abbott Laboratories): composition shown is United States formulation, label claim values.

² Isomil(with Iron (Abbott Laboratories): composition shown is United States formulation, label claim values.

³ Variable, low (negligible) in powders, higher in liquids (see Beach and Menzies, 1986; Hendrickse, Wooldridge and Russell, 1977).

⁴ Maximum inherent level from soy protein isolate ingredient.

⁵ Metabolizable energy using the Atwater conversion factors: protein 17 kJ/g (4 kcal/g), fat 37 kJ/g (9 kcal/g), carbohydrate 17 kJ/g (4 kcal/g).

⁶ Calculated using specific factors from Merrill and Watt (1973: Tables 1 and 8). Carbohydrate in soy protein-based formulas uses a blended figure assuming a mixture of corn syrups and sucrose.

⁷ NME-1: applying NME values to protein, fat and carbohydrate without regard to lactulose or fibre: protein 13 kJ/g (3.2 kcal/g), fat 37 kJ/g (9 kcal/g), carbohydrate 17 kJ/g (4.0 kcal/g).

⁸ NME-2: applying NME values to protein, fat and carbohydrate, but assuming amount of lactose is reduced by a corresponding increase in lactulose, which is calculated as unavailable carbohydrate; energy factor for available carbohydrate assumes present as disaccharides: protein 13 kJ/g (3.2 kcal/g), fat 37 kJ/g (9 kcal/g), lactose 16 kJ/g (3.8 kcal/g).

Note: NME-1 and NME-2 in this table are not the same variables that appear in Figure 3.2 and Table 3.7.

TABLE IV.2A
Effect of using rounded ME and NME conversion values on label claim values for energy of selected baby foods¹

¹ Gerber Products Company, Fremont, Michigan. United States nutrient label claim values (1999).

² Macronutrients for all products are g per 100 g.

³ Label energy values in this column for all products are kJ (kcal) per 100 g declared by the manufacturer.

⁴ Values for kJ for all products are calculated from manufacturer’s values for kcal.

⁵ Total carbohydrate includes values for fibre and sugars.

⁶ Calculated as available carbohydrate (total less fibre) x a factor + fibre x a factor.

⁷ Calculated using the following factors: protein 17 kJ/g (4 kcal/g), fat 37kJ/g (9 kcal/g), available carbohydrate 17 kJ/g (4 kcal/g), fibre 8 kJ/g (1.8 kcal/g).

⁸ Calculated using the following factors: protein 13 kJ/g (3.2 kcal/g), fat 37 kJ/g (9 kcal/g), available carbohydrate 17 kJ/g (4 kcal/g), fibre 6 kJ/g (1.5 kcal/g).

⁹ Taken from rice, dry - granulated for breakfast cereal (item 1882, Table 1) from Merrill and Watt (1973).

¹⁰ Taken from apple sauce, ingredient apples (item 28, Table 1) from Merrill and Watt (1973).

TABLE IV.2B
Effect of using rounded ME and NME conversion values on label claim values for energy of selected baby foods¹

¹ Gerber Products Company, Fremont, Michigan. United States nutrient label claim values (1999).

² Macronutrients for all products are g per 100 g.

³ Label energy values in this column for all products are kJ (kcal) per 100 g declared by the manufacturer.

⁴ Values for kJ for all products are calculated from manufacturer’s values for kcal.

⁵ Total carbohydrate includes values for fibre and sugars.

⁶ Calculated as available carbohydrate (total less fibre) x a factor + fibre x a factor.

⁷ Calculated using the following factors: protein 17 kJ/g (4 kcal/g), fat 37 kJ/g (9 kcal/g), available carbohydrate 17 kJ/g (4 kcal/g), fibre 8 kJ/g (2.0 kcal/g). Atwater specific values for energy are not included because the type of squash used and the proportions of ingredients in the chicken and chicken gravy are not known.

⁸ Calculated using the following factors: protein 13 kJ/g (3.2 kcal/g), fat 37 kJ/g (9 kcal/g), available carbohydrate 17 kJ/g (4 kcal/g), fibre 6 kJ/g (1.5 kcal/g).