In mid-2014, FSANZ commenced a small analytical program to improve the quality and robustness of its food composition data holdings. Foods and beverages were identified for analysis based on current data holdings and frequency of their consumption in the 2011-13 Australian Health Survey (AHS). The analytical program was staged in 4 phases. Phases 1-3 identified 43 of the highest priority foods and beverages for analysis. Phase 4 focused exclusively on alcoholic beverages as the most recent data available to FSANZ for beer and table wines was collated in1990, and the varieties analysed did not accurately reflect current product availability and patterns of consumption.
The range of nutrients analysed differed for each food and beverage depending on what data was available, the quality of the data, whether the nutrient was likely to be present in the food and the impact the consumption of the food or beverage may have on population intakes of that nutrient.
A total of 8 purchases of each food and beverage were made across the Australian Capital Territory, Queensland, Victoria, South Australia and Western Australia to provide a range of brands and production locations. To ensure an appropriate sample weight was obtained, multiple items may have been included in one purchase e.g. one purchase of apples may have included four individual apples. Sampling was carried out by FSANZ and the National Measurement Institute (NMI). All food and beverage purchases were made within capital city metropolitan areas at a retail outlet representing the buying habits of the majority of the community, unless otherwise specified. If more than one sample of the same brand was purchased, different batch codes were selected.
All Phase 4 samples were purchased using a popular national online beer and wine merchant. Due to localised production of beer and wine brands (e.g. all VB is manufactured in Victoria then distributed throughout the country) it was considered that purchasing samples in multiple cities would not improve the representative nature of the sample.
The complete list of foods and beverages included in the 2014-15 key foods program is available in Table 1.
Table 1: Foods and beverages selected for analysis
|Foods and beverages|
No. of samples purchased
(total no. of items purchased)
|No. of brands/varieties|
|Apple, pink lady, unpeeled, fresh/raw||8 (39)||1|
|Mandarin, peeled, fresh/raw||8 (40)||3|
|Orange, peeled, fresh/raw||8 (30)||1|
|Pear, green skinned, unpeeled, fresh/raw||8 (36)||2|
|Baby spinach, fresh/raw||8 (27)||1|
|Rocket, raw fresh/raw||8 (35)||1|
|Carrot, mature, peeled, fresh/raw||8 (82)||1|
|Onion, brown skinned, mature, peeled, fresh/raw||8 (36)||1|
|Coffee, instant, dry powder or granules||8 (17)||7|
|Soft drink, cola flavour, regular||8 ( 8 )||4|
|Tea, green, plain, without milk||8 (10)||7|
|Noodle, buckwheat or soba, boiled, drained||8 (21)||3|
|Noodle, rice stick, boiled, drained||8 (17)||6|
|Noodle, wheat, instant, flavoured, boiled, drained1||8 (75)||6|
|Mango, peeled, raw||8 (25)||2|
|Watermelon, peeled, raw||8 ( 8 )||1|
|Avocado, peeled, raw||8 (31)||1|
|Cucumber, Lebanese, unpeeled, raw||8 (36)||1|
|Tomato, common, raw||8 (44)||2|
|Sauce, tomato, commercial, regular||8 ( 9 )||4|
|Coffee, caffeinated, espresso/short black||8 (37)||8|
|Dairy blend, butter & edible oil spread, regular fat (>70% fat) & salt (>480 mg/100 g Na)||8 (10)||3|
Yoghurt, natural or Greek style, regular fat
(approx. 4% fat)
|8 ( 8 )||6|
|Rice, brown, steamed||8 ( 8 )||4|
|Bacon, short cut, fried in pan, without added fat||8 (10)||5|
|Bacon, middle rasher, fried in pan, without added fat||8 (12)||5|
|Salmon, smoked, as purchased||8 (21)||5|
|Milk, cow, skim, unfortified||8 ( 8 )||8|
|Phase 3|| || |
|Blueberry, fresh, raw ||8 (32)||3|
|Broccoli, fresh, boiled in unsalted water||8 (25)||1|
|Bean, green, fresh, raw ||8 ( 8 )||1|
|Juice, apple, commercial, shelf stable, added vitamin C||8 ( 8 )||5|
|Juice, orange, commercial, fresh & shelf stable, added vitamin C||8 ( 8 )||6|
|Milk, oat, fluid, unfortified||8 ( 8 )||2|
|Milk, cow, fluid, lactose free, regular fat||8 ( 8 )||3|
|Cheese, cheddar, processed, regular fat||8 (11)||5|
|Yoghurt, vanilla flavoured, regular fat (approx. 3% fat)||8 ( 8 )||4|
|Taco shell, from corn flour, plain||8 (27)||6|
|Nut, almond, raw, with skin, unsalted||8 (13)||6|
|Cocoa powder||8 (20)||6|
|Sauce, soy, commercial, regular||8 (10)||6|
|Rice, white, rice cooker||8 ( 8 )||7|
|Rice, white, purchased par-cooked or instant, microwaved||8 (16)||6|
|Beer, high alcohol 5% v/v & above||8 ( 8 )||8|
|Beer, full strength (alcohol 4-4.9% v/v)||8 ( 8 )||8|
|Beer, mid-strength (alcohol 3-3.9% v/v)||8 ( 8 )||8|
|Beer, light (alcohol 1-<3% v/v)||7 ( 7 )||7|
|Beer, mid-strength (alcohol 3 - 4.6% v/v), carbohydrate modified||8 ( 8 )||8|
|Wine, white, dry style (sugar content, 1%), Chardonnay||8 ( 8 )||8|
|Wine, white, dry style (sugar content, 1%), Semillon||8 ( 8 )||8|
|Wine, white, dry style (sugar content, 1%), Sauvignon blanc||8 ( 8 )||8|
|Wine, white, medium dry style, Riesling||8 ( 8 )||8|
|Wine, white, sparkling||8 ( 8 )||8|
|Wine, white, sweet dessert style||8 ( 8 )||8|
|Wine, red, Shiraz ||8 ( 8 )||8|
|Wine, red, Cabernet ||8 ( 8 )||8|
|Wine, red, Merlot||8 ( 8 )||8|
|Wine, red, Pinot noir||8 ( 8 )||8|
|Wine, Rose||8 ( 8 )||8|
|Cider, apple (alcohol ~4-5% v/v)||8 ( 8 )||8|
1. Instant wheat noodle samples were analysed both raw and cooked.
Preparation & analysis
The samples were delivered by hand or sent by courier to NMI. Once received, the samples were photographed and copies provided to FSANZ for approval prior to analysis.
NMI prepared samples according to the sample preparation procedures provided by FSANZ. Each sample was weighed, homogenised and either analysed individually or combined to form one composite sample, depending on the nutrient to be analysed.
NMI conducted the analyses at their Melbourne laboratories using methods of analysis that have been accredited by the National Association of Testing Authorities.
FSANZ validated the results using information from analytical data previously obtained by FSANZ, food labels (ingredient lists and nutrition information panels) where available and international food composition databases.
The majority of results were consistent with previous findings. A small number of analytes in some foods showed levels outside the expected range. These food samples were reanalysed by the laboratory and all results were verified and accepted.
Some notable unexpected results were identified:
- A potassium level of 4400 mg/100 g was recorded for a composite sample of cocoa powder which was substantially higher than previous data obtained by FSANZ. Potassium carbonate is added during the processing of cocoa beans to develop flavour and colour and there may be variation between how much is added and how much remains in the final product. Additional information on cocoa processing can be found on the International Cocoa Organisation website at http://www.icco.org/about-cocoa/processing-cocoa.html.
- Higher than expected levels of iodine were recorded in brown rice, bacon and skim milk samples. The amount of iodine in foods can be affected by several factors. The iodine content in plant foods reflects the iodine content of soil, irrigation water and fertilizers and the iodine content of meat and dairy products may reflect the iodine content of the drinking water, forage and feed and use of iodine feed supplements. Iodised salt may be used in food manufacturing.
- Higher than expected levels of beta-carotene were recorded in broccoli, cow's milk, vanilla yoghurt and cheddar cheese samples. Beta-carotene contributes to colour in plants and its presence in higher levels in milk and cheese than found previously likely reflects feeding differences over time. The higher than expected result in vanilla yoghurt and cheddar cheese could be attributed to the same reason or from the addition of carotene as a colouring agent.
- Total folate levels varied greatly. Levels recorded for pink lady apples, carrots, mandarins, pears and rocket samples were substantially higher than previous data obtained by FSANZ, while levels were lower than expected in broccoli, milk, cheese, cocoa powder and rice samples. The reasons for this are not clear but could reflect natural variation, advances in analytical techniques over the past decade leading to greater extraction of folates from foods, or changes in transport, processing and/or storage that affect post-harvest folate losses.
For some nutrients, the variation between individual samples was considerable (see Table 2).
Table 2: Variation in nutrient values per 100 g in individual food samples
|Analyte||Food||Minimum /100 g||Maximum /100 g|
|copper (mg)||baby spinach||0.083||0.510|
|sodium (mg)||baby spinach||31||240|
This provides further evidence of the extensive natural variation in the nutrient composition of individual fruit and vegetable types. Nutrient levels can be affected by factors such as the variety of the produce item, time of harvest, degree of ripeness, climate, soil conditions including fertiliser application, post-harvest storage and processing and marketing conditions.
Alcohol concentrations in both red and white varieties were higher than data previously obtained by FSANZ in 1990 (see Table 3). As the concentration of alcohol in wine is primarily determined by the amount of sugar in the grapes when harvested, vineyard-related variables including climatic influences, as well as timing of harvest may reflect the changes in alcohol concentrations.
Table 3: Comparison of average alcohol concentrations in white and red wines in 1990 and 2014
Alcohol g/100 g
Alcohol g/100 g
|Wine, white, medium dry style (~ 1% sugars)||9.3||9.7|
|Wine, white, medium sweet style (~ 2.5% sugars)||8.5||9.6|
|Wine, white, sparkling||8.5||9.7|
|Wine, white, sweet dessert style||7.8||8.5|
For the complete set of results generated from this program refer to Appendix 1 Key foods program data table.xls
The results of this analytical program have filled some important data gaps and given FSANZ an improved level of confidence about the composition of commonly consumed foods which contribute significantly to dietary nutrient intakes and foods which are relatively new to the Australian market. The results will also feed into future releases of the FSANZ reference database NUTTAB.