Introduction

What are Nutrient Reference Values?

In the 1991 Recommended Dietary Intakes (RDIRecommended dietary intake) for use in Australia ( NHMRCNational Health and Medical Research Council 1991, Truswell et al 1990) an RDIRecommended dietary intake value, sometimes presented as a range, was developed for each nutrient. The RDIRecommended dietary intake was defined as: "the levels of intake of essential nutrients considered, in the judgement of the NHMRCNational Health and Medical Research Council, on the basis of available scientific knowledge, to be adequate to meet the known nutritional needs of practically all healthy people…they incorporate generous factors to accommodate variations in absorption and metabolism. They therefore apply to group needs. RDIRecommended dietary intakes exceed the actual nutrient requirements of practically all healthy persons and are not synonymous with requirements."

Despite the emphasis on the population basis of the RDIRecommended dietary intake, the RDIRecommended dietary intakes were often misused in assessing dietary adequacy of individuals, or even foods, not only in Australia and New Zealand but also in many other countries. To overcome this misuse, many countries have moved to a system of reference values that retains the concept of the RDIRecommended dietary intake while attempting to identify the average requirements needed by individuals. In 1991, the UK (Dept Health 1991) became the first country to develop a set of values for each nutrient. More recently, the Food and Nutrition Board: Institute of Medicine (FNB:IOMFood and Nutrition Board: Institute of Medicine 1997, 1998a, 2000a, 2001, 2002, 2004) adopted a similar approach on behalf of the US and Canadian Governments.

After due consideration, the Working Party decided to adopt the approach of the US:Canadian Dietary Reference Intakes (DRIDietary reference intakess) but vary some of the terminology, notably to retain the term 'Recommended Dietary Intake'.

For each nutrient, an Estimated Average Requirement (EAREstimated average requirement) was set from which an RDIRecommended dietary intake could be derived. (Note that the US:Canadian terminology is 'Recommended Dietary Allowance', or 'RDA'). Whilst the various NRVNutrient Reference Valuess are expressed on a per day basis, they should apply to intakes assessed over a period of about 3 to 4 days. If the standard deviation (SD) of the EAREstimated average requirement is available and the requirement for the nutrient is symmetrically distributed, the RDIRecommended dietary intake is set at 2SD above the EAREstimated average requirement. Such that

RDI = EAR +2SD_EAR.

If data about variability in requirements are insufficient to calculate an SDStandard deviation (which is usually the case), a coefficient of variation (CVCoefficient of variation) is used. A CVCoefficient of variation of 10% for the EAREstimated average requirement is assumed for nutrients unless available data indicate that greater variation is probable. The 10% is based on extensive data on variation in basal metabolic rate and protein requirements (FAOFood and Agricultural Organization of the United Nations:WHOWorld Health Organization of the United Nations:UNA 1985, Garby & Lammert 1984, Elia 1992).

If 10% is assumed to be the CVCoefficient of variation, then twice that amount added to the EAREstimated average requirement is defined as equal to the RDIRecommended dietary intake. Thus for a CVCoefficient of variation of 10%, the RDIRecommended dietary intake would be 1.2 x EAREstimated average requirement; for a CVCoefficient of variation of 15% it would be 1.3 x EAREstimated average requirement and for a CVCoefficient of variation of 20% it would be 1.4 x EAREstimated average requirement.

Where evidence was insufficient or too conflicting to establish an EAREstimated average requirement (and thus an RDIRecommended dietary intake) an Adequate Intake (AIAdequate intake) was set, either on experimental evidence or by adopting the most recently available population median intake and assuming that the Australian/New Zealand populations were not deficient for that particular nutrient. Both the RDIRecommended dietary intake and AIAdequate intake can be used as a goal for individual intake, but there is less certainty about the AIAdequate intake value as it depends to a greater degree on judgement. An AIAdequate intake might deviate significantly from and be numerically higher than an RDIRecommended dietary intake if the RDIRecommended dietary intake could be determined. Thus AIAdequate intakes should be interpreted with greater caution.

Where AIAdequate intakes were based on median population intakes, these were derived from a re-analysis of the complete databases of the National Nutrition Surveys of Australia, 1995 (Australian Bureau of Statistics 1998) and New Zealand 1991, 1997, 2002 (LINZLife in New Zealand Activity and Health Research Unit 1992, Ministry of Health 1999, 2003) using the appropriate age bands. The two-day adjusted data were used for the estimates.

For infants of 0 to 6 months, all recommendations are in the form of Adequate Intakes based on the composition of breast milk from healthy mothers, using a standard milk volume. The bioavailability of nutrients in formulas may vary from that in breast milk, so formula-fed babies may need higher nutrient intakes. As formulas can vary in the chemical form and source of the nutrients, it is not possible to develop a single reference value for all formula-fed infants.

For energy, an Estimated Energy Requirement (EER) was set for a range of activity levels for individuals of a specified age, gender and body size.

For each nutrient, an Upper Level of Intake (ULUpper level of intake) was set, which, unless otherwise stated, includes intake from all sources including foods, nutrients added to foods, pills, capsules or medicines. The ULUpper level of intake is the highest average daily nutrient intake level likely to pose no adverse health effects to almost all individuals in the general population. In setting the ULUpper level of intake, any adverse health effect were considered, including those on chronic disease status. The ULUpper level of intake is not a recommended level of intake. It is based on a risk assessment of nutrients that involves establishment of a No Adverse Effect Level (NOAELNo observed adverse effect level) and/or a Lowest Adverse Effect Level (LOAELLowest observed adverse effect level) and application of an Uncertainty Factor (UF) related to the evidence base and severity of potential adverse effects. Members of the general population should be advised not to routinely exceed the ULUpper level of intake. Intakes above the ULUpper level of intake may be appropriate for some nutrients for investigation in well-controlled clinical trials as long as signed informed consent is given and as long as the trials employ appropriate safety monitoring of trial subjects. Readers are referred to the relevant FNB:IOMFood and Nutrition Board: Institute of Medicine documents and the report of the UK Expert Group on Vitamins and Minerals (2003) for more details about the potential toxicological effects of high intakes of nutrients. In Australia, vitamin and mineral supplements are regulated under the Therapeutic Goods Act (1989) that also sets some standards for these products. In New Zealand, dietary supplements are generally regulated under the New Zealand Dietary Supplements Regulations (New Zealand Government 1985), but supplements with nutrients at higher/pharmacological doses than the specified maximum daily doses need to meet the requirements of the Medicines Regulations (1984).

Further details of the approach used in setting ULUpper level of intakes are given in the FNB:IOMFood and Nutrition Board: Institute of Medicine publication Dietary Reference Intakes. A risk assessment model for establishing upper intake levels for nutrients (1998b) and in the relevant nutrient chapters of the DRIDietary reference intakes publications.

The uses of the various NRVNutrient Reference Valuess are summarised in the table below that was adapted from the FNB:IOMFood and Nutrition Board: Institute of Medicine (2000b) publication, Dietary Reference Intakes. Applications in Dietary Assessment. This document also provides further details of potential applications.

In contrast to the US:Canadian approach, the committee agreed to retain the traditional concept of adequate physiological or metabolic function and/or avoidance of deficiency states as the prime reference point for establishing the EAREstimated average requirement and RDIRecommended dietary intakes and to deal separately with the issue of chronic disease prevention. It was felt that assessing nutrient needs for chronic disease prevention in a quantitative manner was still problematical. Research findings related to chronic disease prevention often relate to nutrient mixes or food intake patterns, rather than the intake of individual nutrients.

To address the issue of chronic disease prevention, two additional sets of reference values were developed for selected nutrients for which sufficient evidence existed. The set dealing with the macronutrients was adapted from the work of the FNB:IOMFood and Nutrition Board: Institute of MedicineDRIDietary reference intakes review of macronutrients (2002) and is called the Acceptable Macronutrient Distribution Range (AMDR). The second set of reference values was termed Suggested Dietary Targets (SDTSuggested dietary targets). These related to nutrients for which there was a reasonable body of evidence of a potential chronic disease preventive effect at levels substantially higher than the EAREstimated average requirement and RDIRecommended dietary intake or AIAdequate intake. As the evidence base for chronic disease prevention is mainly derived from studies and health outcomes in adults, these AMDRs and SDTSuggested dietary targets apply only to adults and adolescents of 14 years and over.

The nutrients reviewed

Having considered emerging evidence on the connections between diet and health and the recent recommendations from other countries, the preliminary workshops identified more than 40 nutrients for the Working Party to consider. The document Recommended Dietary Intakes for use in Australia ( NHMRCNational Health and Medical Research Council 1991), which had also been adopted for use in New Zealand, contained recommendations for 19 nutrients and dietary energy. During this review, dietary energy requirements and requirements for the nutrients were considered. Those for which values were set are listed below:

In addition to the nutrients listed above, we also reviewed the literature on total fat (for ages and life stages other than infancy), carbohydrate (for ages and life stages other than infancy), cholesterol, arsenic, boron, nickel, silicon and vanadium. For these nutrients or age bands and life stages, it was agreed that there was little or no evidence for their essentiality in humans. This was generally in line with the findings of the US:Canadian DRIDietary reference intakes review recommendations. However, the DRIDietary reference intakes reviews set upper limits for some of these nutrients (FNB:IOMFood and Nutrition Board: Institute of Medicine 1998, 2001) and the reader is referred to these for information.

The reviews were based on assessment of the applicability of the recently developed US:Canadian Dietary Reference Intakes (FNB:IOMFood and Nutrition Board: Institute of Medicine 1997, 1998a,b, 2000a,b, 2001, 2002, 2004) to Australia and New Zealand, with reference to recommendations from other countries such as the UK (1991, 2003), Germany:Austria:Switzerland (DACH recommendations 2002) and from key organisations such as the FAOFood and Agricultural Organization of the United Nations:WHOWorld Health Organization of the United Nations (2001).

Reference body weights

In developing the recommendations it was necessary to standardise body weights for the various age/gender groups. Assessment of the data on measured body weights and heights for relevant age/gender categories from the most recent National Nutrition Survey of Australia, 1995 (ABS 1998) and New Zealand, 1997 and 2002 (MOHMinistry of Health 1999, 2003) showed that the body weights were similar to those used in the earlier US:Canadian DRIDietary reference intakes publications. From the 2002 publication onwards, the US:Canadian DRIDietary reference intakes review panels changed their standard body weights in response to availability of new data showing markedly lighter body weights than previously used. As the most recent Australian/New Zealand data more closely resembled those in the earlier US:Canadian reports, these were adopted for use throughout these recommendations.

The standard body weights for all adults were based on that for 19–30 year olds, although body weight in most western populations tends to increase throughout adulthood because of increasing body fat.

Extrapolation processes

Experimental data are often only available for a limited age/gender group. The setting of recommendations for other groups may require extrapolation of the data. This is sometimes based on energy requirements, but more commonly on a metabolic body weight. In extrapolating data from one group to another, the processes and formulae used were those developed by the US:Canadian DRIDietary reference intakes panels unless otherwise indicated in the text.

Extrapolations from adult Estimated Average Requirements (EAREstimated average requirement) to children’s requirements were mostly done using the formula:

The growth factors used were 0.3 from 7 months to 3 years of age and 0.15 for 4–13 years of age for both genders. For boys aged 14–18 years, the growth factor used was 0.15 but for girls of this age, the growth factor was set at zero.

When extrapolating from the Adequate Intake (AIAdequate intake) for younger infants aged 0-6 months, to older infants aged 7-12 months, the formula used was:

When estimating the Upper Level of Intake for children, the ULUpper level of intake was extrapolated down from the adults ULUpper level of intake using the formula:

This allows both body mass and metabolic differences between adults and children to be incorporated as necessary. For more details can be found in the methodology sections of the United States:Canadian FNB:IOMFood and Nutrition Board: Institute of Medicine reports.

Implications

The implications for adoption of these revised NRVNutrient Reference Valuess include:

• The need to address ongoing education of both health and food industry professionals in the end use of the various reference values and related tools for their use.

• The need to update a number of documents and educational tools based on the previous RDIRecommended dietary intakes, including:

  • The NHMRCNational Health and Medical Research Council Core Food Groups Analysis ( NHMRCNational Health and Medical Research Council 1994)

  • The Australian Guide to Healthy Eating and the Dietary Guidelines for Australian Adults, the Dietary Guidelines for Children and Adolescents in Australia and the Dietary Guidelines for Older Australians

  • The New Zealand Food and Nutrition Guidelines for the ages and stages of the lifecycle.

  In Australia, the Core Food Group analysis addressed the translation of the nutrient recommendations into amounts of core foods (eg cereals, fruits and vegetables, meats, fish, poultry, dairy, fats and oils) required to meet these nutrient recommendations in Australia. These in turn were used as the basis for the development of the Australian Guide to Healthy Eating and the Australian Dietary Guidelines for Adults, the Dietary Guidelines for Children and Adolescents in Australia and the Dietary Guidelines for Older Australians.

  New Zealand has Food and Nutrition Guidelines covering the ages and stages of the lifecycle. There are currently seven in the series including infants and toddlers (0–2 years), children (2–12 years), adolescents, pregnant women, breastfeeding women, adults and older people. These publications include a background paper for health professionals and an accompanying health education pamphlet for the public.

  The interrelationships between these various recommendations and the underpinning evidence are shown in Figure 1.

• The need for regular monitoring of dietary intake and nutrient status in the population, including the use of fortified foods and supplements, to underpin the ongoing revisions of the NRVNutrient Reference Valuess, notably the Adequate Intake values which, by definition, are often based on population median dietary intakes.

• The need for research funds to enable more accurate assessment of requirements for both sustenance and prevention of chronic disease, including studies on issues such as biomarkers for nutritional status and nutrient bioavailability, and adverse effects of high intakes.

• The need to update and expand existing food databases for the analysis of national nutrition survey data, including information on the levels of fortification in foods.

• The need to change computerised dietary analysis programs that use the existing RDIRecommended dietary intake values as reference values.

• The need for the redevelopment of relevant standards for the use of NRVNutrient Reference Valuess for food legislative purposes, including issues such as food labelling and food fortification.

• The need to consider the implications of changes in the NRVNutrient Reference Valuess for the food and dietary supplementation industry.

Figure 1. Interrelationships between the evidence base, NRVNutrient Reference Valuess, Core Food Group Analysis, Dietary and Food Guidelines and Healthy Eating Guides

What are the implications of changes in recommendations for certain nutrients?

Consumption of a diet conforming to the NRVNutrient Reference Valuess need not, in itself, be more expensive for the individual (Baghurst 2003), however addressing the needs for implementation outlined above will involve ongoing costs that are difficult to quantify. The financial expense associated with inadequate nutrition in the community is likely to far outweigh that of implementing the necessary changes. Crowley et al (1992) have estimated the economic cost of diet-related disease in Australia in terms of both direct health care (hospitals, medical expenses, allied health professional services, pharmaceutical expenses and nursing homes) attributable to diet and indirect costs (due to sick leave and the net present value of forgone earnings due to premature death). The estimate of direct costs, excluding consideration of alcohol, was $1,432 million and that for indirect, $605 million, giving a total of $2,037 million for 1989–1990.

The RDIRecommended dietary intake for some nutrients has substantially increased from that in the previous edition due to the availability of new data or changes in the way needs are assessed. In the past, needs at the individual level were often assessed in the practical situation by reference to 70% RDIRecommended dietary intake in the absence of a specific EAREstimated average requirement value. The NHMRCNational Health and Medical Research Council Core Food Group assessment, which is the basis for the Australian Guide to Healthy Eating, was also modelled on 70% RDI