Long-chain n-3 PUFA are associated with several health benefits in adults and children. There is growing evidence that long-chain n-3 PUFA may be beneficial in the treatment of several childhood and adolescent disorders such as asthma, depression and bipolar disorder(Reference Clayton, Hanstock and Garg1).
The National Health and Medical Research Council nutrient reference values for recommended intakes of long-chain n-3 PUFA in Australia and New Zealand(2) include EPA and DHA, which are found in high concentrations in fish, as well as docosapentaenoic acid, which is also found in relatively high concentrations in meat other than fish. Meat and, in particular, lean red meat including beef, veal and lamb, contributes significantly to total long-chain n-3 PUFA intake in Australian adults(Reference Howe, Meyer and Record3); however, there are limited published data available examining the contribution of meat to long-chain n-3 PUFA intake in Australian children and adolescents.
In order to estimate the contribution of meat to long-chain n-3 PUFA intake, it is necessary to know the intake of individual meat sources. Several studies have measured the intake of meat by Australian children and adolescents; however, the intake of individual meat sources is often not reported. The aim of the present review was to examine published intakes of meat and fish by children and adolescents in Australia and to estimate the intake of individual meat sources compared with recommended intakes.
A review of studies and surveys containing published values for meat or fish intake in Australia was conducted using PubMed, Agricola and CAB Abstracts using the terms ‘intake’, ‘beef’, ‘lamb’, ‘pork’, ‘poultry’, ‘fish’, ‘children’ and ‘adolescents’ and using published reference lists. Internet searches were also conducted using web search engines such as Google™, Yahoo!Search®, Live Search and Ask™.
Reported intakes of meat and fish
Two nationwide surveys of child and adolescent intakes were conducted in 1985 and 1995. Meat intake was also reported in two regional studies; one in Western Sydney in 1994 and one in Western Australia in 2003.
National Dietary Survey of Schoolchildren aged 10–15 years (1985)
The first national survey of dietary intake was the National Dietary Survey of Schoolchildren (NDSSC)(Reference English, Cashel and Lewis4). School children aged 10–15 years across all Australian states and territories (n 5224) completed a 1 d food record between May and October 1985(Reference English, Cashel and Lewis4). Trained staff instructed students on how to complete food records and provided standard cup and spoon measures, rulers and diagrams to assist in estimating portion size.
Weighted average intakes of individual meat sources including mixed dishes (not including casserole gravy)(Reference English, Cashel and Lewis4) are reported in Table 1. Median intakes for all participants were presented in a separate publication(Reference Cook, Rutishauser and Allsopp5) and were significantly lower than mean values. The median intake of total meat was 104 and 80 g/d for males and females, respectively, while the median intake of fish (fish and seafood) was 0 g/d for both males and females.
NDSSC, National Dietary Survey of School children; NNS, National Nutrition Survey
* Beef = beef + veal; lamb = lamb + mutton; total meat other than fish = beef + lamb + pork + poultry including mixed dishes; fish = fish + seafood.
† Total meat other than fish includes mixed dishes without casserole gravy but not take-away dishes.
‡ Number of participants in each age range calculated from the percentage of males and females in the overall sample.
National Nutrition Survey (1995)
As part of the 1995 National Nutrition Survey (NNS) the intake of children aged 8–18 years was estimated using a 24 h dietary recall interview(Reference McLennan and Podger6). Children aged 8–11 years undertook the interview with a trained nutritionist with the assistance of a parent or guardian; participants over the age of 11 years were interviewed without parental assistance. Mean intakes of individual meat sources were calculated from published values (see below). Again, median intakes were lower than mean values(Reference Cook, Rutishauser and Allsopp5). The median intake of total meat was 111 and 78 g/d for males and females, respectively, and the median intake of fish was 0 g/d for both males and females(Reference Cook, Rutishauser and Allsopp5).
Calculation of intake of individual meat sources from the 1995 National Nutrition Survey
The original publication from the 1995 NNS(Reference McLennan and Podger6) did not present data for the intake of individual meat sources. Intakes were calculated from original data and subsequent reports of red meat consumption(Reference Baghurst7).
Total meat is defined as beef + veal + lamb + pork + bacon + ham + poultry including mixed dishes(Reference McLennan and Podger6), while red meat is defined as beef + veal + lamb including mixed dishes(Reference Baghurst7). The intake of total meat other than poultry was calculated from:
Intake of total meat other than poultry = intake of total meat − intake of poultry and mixed dishes (where poultry is the major component)(Reference McLennan and Podger6).
The intake of pork, ham and bacon was then calculated as:
Intake of pork + ham + bacon = intake of total meat other than poultry − intake of red meat(Reference Baghurst7).
Finally, from the reported intake of red meat(Reference Baghurst7), the intake of beef + veal or lamb was calculated using the proportion of red meat consumed as beef + veal reported in a similar population(Reference Laing, Oram and Burgess8):
Intake of beef + veal = intake of red meat(Reference Baghurst7) × (beef + veal) as a percentage of total red meat intake.
For example, the intake of beef + veal was 65·4 % of total red meat intake for males aged 8–11 years.
Survey of School Children in Western Sydney (1993–5)
The meat intake of 999 children aged 1–16 years in Western Sydney was estimated using 4 d food records between 1993 and 1995(Reference Laing, Oram and Burgess8). Instructions for completing food records were provided by teachers and standard cup measures and rulers were provided to assist in portion size estimation. The total intake of beef, lamb, pork or poultry (g) over the 4 d recording period was presented only for those participants who reported consuming each meat source.
The authors could not be contacted to determine the percentage of participants who consumed each meat type, so an average consumption for all participants was calculated using an estimate of the percentage of participants consuming each meat type(Reference McLennan and Podger6) (JF Watson, unpublished results). Values used were: beef, 77 %; lamb, 72 %; poultry, 77 %; pork, 70 %.
Survey of School Children in Western Australia (2003)
The Child and Adolescent Physical Activity and Nutrition Survey (CAPANS) estimated the intake of 1494 school students aged 8–15 years in Western Australia. Intake was estimated using a 1 d food record with a similar to protocol to the 1985 NDSSC(Reference Glasson, Read and Hands9). Students not completing the food record undertook a 24 h recall. Beef, lamb or pork intake was calculated from reported intakes of total meat other than fish and poultry(Reference Glasson, Read and Hands9) using the proportions of each meat type reported in a similar population(Reference Laing, Oram and Burgess8) (Table 1).
The Australian Guide to Healthy Eating (AGHE)(Reference Smith, Kellett and Schmerlaib10, 11) recommends children aged 8–11 years consume 1–1·5 serves per d of lean meat, fish or poultry, while it is recommended that children aged 12–18 years consume 1–2 serves per d. The AGHE reports a serve of cooked beef, lamb, pork or chicken is 65–100 g and cooked fish is 80–120 g(Reference Smith, Kellett and Schmerlaib10). Standard serves of meat are also reportedly 135 g(Reference Williams12).
The lower recommended intake of red meat for children aged 12–18 years calculated from the AGHE is one serve per d, three times per week at 65 g per serve, or approximately 27·9 g/d (Fig. 1). The upper recommended intake for these children is approximately 114·3 g/d (two serves per d; four times per week; 100 g per serve). Recommended intakes of fish were also calculated assuming fish is consumed one or two times per week.
Summary of intake data and comparison with recommended intakes
Weighted mean intakes of red meat and fish were calculated for studies conducted between 1994 and 2003 and compared with recommended intakes (Fig. 1).
Estimated intakes of meat and fish by children and adolescents in Australia are available, in part, from four key studies. Differences in methods, including study population, number of days collected, portion size estimation and whether trained staff or teachers assisted with data collection, age ranges, categories of meat consumption and details of data reporting make it difficult to compare meat intake between studies. For example, the 1995 NNS is strengthened by the breadth of sample population across Australia compared with the regional sample in Western Sydney; however, it is weakened by the use of 24 h recalls compared with 4 d food records. Several assumptions about data and populations were required to allow comparison. Some limitations of the surveys conducted and calculations performed are briefly reviewed below.
Comparison of methodology
The studies reviewed used food records or 24 h recalls. Dietary intakes are underestimated using both methods, with greater under-reporting with increasing age(Reference Bandini, Cyr and Must13). Energy intake may be under-reported by as much as 20 % using food records and by a further 16 % using 24 h recalls(Reference Buzzard and Willet14). One study(Reference Glasson, Read and Hands9) used a combination of 24 h recalls and food records, but did not report the proportion of methods used. Therefore, estimated intakes of total meat and fish (Table 1) are likely to underestimate actual intakes, particularly for adolescents.
Although food records and 24 h recalls capture rich information on food consumption, episodically consumed foods are not accurately measured(Reference Subar, Dodd and Guenther15) and may be underestimated in the studies reviewed. Although the Western Sydney study(Reference Laing, Oram and Burgess8) collected multiple records, biases for variation and under-reporting may still remain(Reference Black and Cole16).
With the exception of 8- to 11-year-olds in the 1995 NNS, intakes for the studies reviewed were self-reported by participants. Children aged 8 years or older can reliably report their food intake(Reference Riley17) and the use of age-appropriate food photographs(Reference Foster, Matthews and Nelson18) or portion size software(Reference Foster, Matthews and Lloyd19) can improve the ability of children and adolescents to estimate portion size. However, children and adolescents may not accurately quantify portion sizes with standard household measures and diagrams used in the studies reviewed(Reference Livingstone and Robson20). Individuals who eat larger portions may underestimate intake while those consuming smaller portions overestimate(Reference Biró, Hulshof and Ovesen21). It is not possible to accurately determine the impact of portion size estimation on the intakes reported.
Differences between study populations may also influence reported intakes. For example, the prevalence of overweight and obese children and adolescents doubled to 25 % in New South Wales between 1985 and 1997(Reference Booth, Chey and Wake22) and participant BMI was higher in the 1995 NNS compared with the 1985 NDSSC. Overweight children and adolescents under-report energy intake by up to 40 % compared with their non-overweight peers(Reference Domel23).
Estimation of errors in the calculation of meat intake
Calculated intakes of beef or lamb for the 1995 NNS are affected by the estimate of the proportion of red meat consumed as beef or lamb. Every 1 % overestimation of this proportion overestimates intake by approximately 1 % of reported red meat intake. For example, beef intake by males aged 16–18 years is overestimated by 0·82 g/d (1 % of 82 g/d) for every 1 % overestimation of beef as a proportion of red meat. Similarly, in the Western Australia study(Reference Glasson, Read and Hands9), beef, lamb or pork intake is over- or underestimated by 1 % of total meat intake (other than poultry or fish) for every 1 % over- or underestimation of the proportion of each meat source.
In the Western Sydney study(Reference Laing, Oram and Burgess8), every 1 % overestimation of participants consuming each meat type overestimates calculated intake by approximately 1·2–1·4 %. For example, the intake of lamb by males aged 14–16 years is overestimated by 0·53 g/d (1·4 % of 37·9 g) if only 71 % instead of 72 % of participants consumed lamb.
Intakes of total fish and seafood were reported directly in three studies reviewed. However, intakes of different types of fish in these age ranges are not published. Long-chain n-3 PUFA concentrations are significantly higher in oily fish compared with non-oily fish(Reference Sinclair, Oon and Lim24). Therefore, the contribution of fish to total long-chain n-3 PUFA intake cannot easily be determined.
There are currently no published data available examining the contribution of meat (other than fish) to long-chain n-3 PUFA intake in children and adolescents. However, as the intake of meat (Table 1) is similar to values reported in adults(Reference McLennan and Podger6), meat may contribute significantly to total long-chain n-3 PUFA intake in young individuals in Australia similar to adults(Reference Howe, Meyer and Record3). Similarly to fish, knowledge of the intake of individual meat cuts is also required in order to accurately estimate the contribution of meat to total long-chain n-3 PUFA intake.
Reported v. recommended intakes
Weighted mean intakes of red meat in the three most recent studies were within recommended intakes (Fig. 1). Similarly, reported intakes of fish meet target guidelines, although intakes were towards the lower level even if fish is recommended only once per week. While the intake of red meat and fish appear to be adequate in the age ranges examined, it is important to consider that the recommended ranges are very large and reported median intakes are also lower than mean intakes(Reference Cook, Rutishauser and Allsopp5).
Intakes of meat generally meet recommended guidelines for children and adolescents in Australia and may contribute significantly to total long-chain n-3 PUFA intake. Although there are several limitations to the estimations and calculations in the present review, many of these errors are small and may be lower than errors associated with the data collection methods themselves. Given changes in the prevalence of overweight and obesity, for example, there is a great need for a more updated survey of food consumption in children and adolescents in Australia.
We thank Dr Trent Watson, Dr Robert Taylor and Dr John Wilkins for valuable contributions in drafting and editing of the manuscript. This review was supported from in-kind contributions from the Hunter New England Area Health Service. The authors all contributed to the production of the manuscript. All contributed to the review of literature, data interpretation and drafting and review of the manuscript.
The authors of the present study do not have any commercial associations that might pose a conflict of interest in connection with this paper.