25th Australian Total Diet Study

Food Standards Australia New Zealand

FSANZ 2019 - 25th Australian Total Diet Study

Food Standards Australia New Zealand’s 25th Australian Total Diet Study measured agricultural/veterinary chemicals and metal contaminants in a broad Australian food basket. For metals, FSANZ analysed 508 prepared/as-consumed composite samples representing 88 food types purchased across all Australian states and territories in two sampling windows (13-24 May 2013 and 17-28 February 2014). The metals chapter reports total arsenic, inorganic arsenic, cadmium, lead, total mercury, inorganic mercury, and methylmercury occurrence and dietary-exposure estimates, then compares relevant foods with Australia New Zealand Food Standards Code maximum levels.

Key numbers

Total arsenic (tAs):

Detected in 45 of 88 sampled foods; 142 of 508 composite samples (28%) had detectable total arsenic and 366 composites (72%) had no detectable total arsenic (p. 32).
Highest mean concentrations, with non-detects set equal to the LOR (nd=LOR), were prawns 2.9 mg/kg, mussels 2.7 mg/kg, takeaway plain fish fillets 2.2 mg/kg, canned tuna in brine 0.92 mg/kg, and frozen fish portions 0.88 mg/kg; the highest non-seafood mean was rice-based breakfast cereal at 0.26 mg/kg (p. 33; Table 5).
No composite sample exceeded its corresponding Schedule 19 / Standard 1.4.1 commodity maximum level for total arsenic (p. 33).
Total-arsenic dietary exposures for consumers were 0.37-0.90 µg/kg bw/day mean and 1.3-2.8 µg/kg bw/day P90 at the lower bound (nd=0), and 0.49-1.2 µg/kg bw/day mean and 1.4-3.1 µg/kg bw/day P90 at the upper bound (nd=LOR). Children aged 2-5 years had the highest mean and P90 exposure on a body-weight basis (p. 34).
Major total-arsenic contributors were meat/poultry/seafood/eggs (66-87%) and cereals/cereal products (8-20%). Within the seafood category, plain fish contributed 25-39%, crumbed/battered fish and seafood 13-30%, tuna/all canned and smoked seafood 7-17%, and crustacea 5-16% except for children aged 2-5 years (pp. 35-36).

Inorganic arsenic (iAs):

Analysed only in likely contributor foods: takeaway fish fillets, frozen fish portions, mussels, prawns, rice, sushi, and canned tuna in brine (Table 1, p. 17).
Detected in three foods at mean concentrations (nd=LOR): mussels 0.28 mg/kg, white rice 0.03 mg/kg, and sushi rolls/nori 0.01 mg/kg. Takeaway fish fillets, frozen fish portions, prawns, and canned tuna had no detectable inorganic arsenic (p. 36).
No composite sample exceeded corresponding inorganic-arsenic maximum levels (p. 36).
Using the analysed-samples-only method, consumer inorganic-arsenic exposures were 0.054-0.10 µg/kg bw/day mean and 0.12-0.26 µg/kg bw/day P90 at the lower bound (nd=0), and 0.042-0.092 µg/kg bw/day mean and 0.10-0.23 µg/kg bw/day P90 at the upper bound (nd=LOR). For infants aged 9 months, respondent mean/P90 exposures were 0.020/0.040 µg/kg bw/day lower-bound and 0.023/0.046 µg/kg bw/day upper-bound (pp. 36-37).
Food-category contribution for iAs was dominated by cereals/cereal products, specifically rice and rice products (97-100%); molluscs contributed <1-3% (p. 38).

Cadmium (Cd):

Detected in 43 of 88 foods; 149 of 508 composite samples (29%) had detectable cadmium and 359 (71%) did not (p. 40).
Fourteen foods had cadmium detected in all of their respective composites: mussels, potato chips, potato, peanut butter, desiccated coconut, garlic, chocolate cake, chocolate, multigrain bread, canned beetroot, canned tuna, carrots, canned tomatoes, and pizza (p. 40).
Highest mean cadmium concentrations (nd=LOR) were mussels 0.20 mg/kg, prawns 0.065 mg/kg, potato crisps 0.058 mg/kg, and potato 0.026 mg/kg; all other food means were below 0.020 mg/kg (pp. 40-41). Table 6 reports canned tuna at 0.012 mg/kg (p. 42).
No composite sample exceeded its corresponding Schedule 19 / Standard 1.4.1 maximum level for cadmium (p. 41).
Consumer cadmium dietary exposure was 2.0-5.5 µg/kg bw/month mean and 3.7-9.9 µg/kg bw/month P90 at the lower bound (nd=0), and 5.8-14 µg/kg bw/month mean at the upper bound (nd=LOR). The source text prints the upper-bound P90 range as 8.8-20 µg/kg bw/day inside this monthly PTMI paragraph; because that unit conflicts with the surrounding cadmium/month context, the page flags the value rather than normalizing it (p. 42).
Cadmium exposures were below the PTMI for all assessed groups except a slight upper-bound P90 exceedance for 9-month infants (130% of PTMI), which FSANZ characterized as short-term and conservative (pp. 44-45).

Lead (Pb):

Detected in 36 of 88 foods; 76 of 508 composite samples (15%) had detectable lead and 432 (85%) did not (p. 46).
Highest mean lead concentrations (nd=LOR) were mussels 0.074 mg/kg, sultanas 0.037 mg/kg, chocolate cake 0.026 mg/kg, and honey 0.024 mg/kg; all other foods were below 0.020 mg/kg (pp. 46-47).
No composite sample exceeded its corresponding Schedule 19 / Standard 1.4.1 maximum level for lead (p. 47).
Consumer lead dietary exposure was 0.016-0.048 µg/kg bw/day mean and 0.032-0.10 µg/kg bw/day P90 at the lower bound (nd=0), and 0.16-0.38 µg/kg bw/day mean and 0.23-0.56 µg/kg bw/day P90 at the upper bound (nd=LOR). Infants aged 9 months had respondent mean/P90 exposures of 0.040/0.079 µg/kg bw/day lower-bound and 0.51/1.0 µg/kg bw/day upper-bound (p. 48).
Major food-category contributors were beverages (28-57%), fruits/nuts (14-36%), cereals/cereal products (9-29%), meat/poultry/seafood/eggs (8-10% for ages 6+), and sugars/confectionery (6-9%). Within meat/poultry/seafood/eggs, pork and deli meats contributed 6-7%; crustacea (<1%) and molluscs (<1-4%) were minor contributors (pp. 49-50).
Table 8 reports MOEs for ages 2-5 years of 6 (mean nd=0), 0.8 (mean nd=LOR), 3 (P90 nd=0), and 0.5 (P90 nd=LOR), with higher MOEs in older groups (p. 52).

Total mercury (tHg):

Detected in 14 of 88 foods; 40 of 508 composite samples (8%) had detectable mercury and 468 (92%) did not. Six mercury-detect foods were seafoods (p. 53).
Highest mean total-mercury concentrations (nd=LOR) were takeaway fish fillets 0.13 mg/kg, frozen fish portions 0.048 mg/kg, and canned tuna in brine 0.046 mg/kg. Table 9 also reports mussels 0.010 mg/kg and prawns 0.015 mg/kg (pp. 53-54).
No composite sample exceeded Schedule 19 / Standard 1.4.1 mercury maximum levels (p. 54).
Consumer total-mercury dietary exposure was 0.32-1.1 µg/kg bw/week mean and 0.73-2.4 µg/kg bw/week P90 at the lower bound (nd=0), and 1.1-2.7 µg/kg bw/week mean and 1.7-4.1 µg/kg bw/week P90 at the upper bound (nd=LOR). For infants aged 9 months, respondent mean/P90 exposures were 0.15/0.29 µg/kg bw/week lower-bound and 3.6/7.1 µg/kg bw/week upper-bound (pp. 54-55).
Meat/poultry/seafood/eggs contributed 98-100% of total-mercury exposure; within that category, plain fish contributed 51-71%, crumbed/battered fish and seafood 15-34%, and tuna/all canned and smoked seafood 11-26% (p. 55).

Inorganic mercury (iHg):

Analysed in selected seafoods with a focus on likely dietary exposure sources; there were no inorganic-mercury detections in those foods (p. 56).
Consumer inorganic-mercury dietary exposure was 0.055-0.19 µg/kg bw/week mean and 0.09-0.32 µg/kg bw/week P90 at the lower bound (nd=0), and 1.0-2.4 µg/kg bw/week mean and 1.5-3.6 µg/kg bw/week P90 at the upper bound (nd=LOR). For 9-month infants, respondent exposures were 0.00003/0.000061 µg/kg bw/week mean/P90 lower-bound and 3.4/6.8 µg/kg bw/week mean/P90 upper-bound (pp. 56-57).
The sole food-category contributor in the model was takeaway foods/snacks, entirely from sushi roll, because sushi was the only food with detectable total mercury that FSANZ assumed to be inorganic mercury for this assessment (p. 57).

Methylmercury (MeHg):

Analysed in selected seafoods; detected in three foods at mean concentrations (nd=LOR): takeaway fish fillets 0.14 mg/kg, frozen fish portions 0.06 mg/kg, and canned tuna 0.05 mg/kg. Mussels and prawns had no detectable methylmercury (p. 59).
Consumer methylmercury dietary exposure was 0.55-1.7 µg/kg bw/week mean and 1.2-3.5 µg/kg bw/week P90 at the lower bound (nd=0), and 0.38-1.2 µg/kg bw/week mean and 0.8-2.4 µg/kg bw/week P90 at the upper bound (nd=LOR). The lower-bound range is higher than the upper-bound range because the two scenarios have different consumer counts (p. 59).
Females aged 16-44 years, used by FSANZ as a proxy group for pregnant women, had 0.61 µg/kg bw/week mean and 1.4 µg/kg bw/week P90 lower-bound exposure; upper-bound consumer mean/P90 exposures were 0.43/0.87 µg/kg bw/week (p. 59).
Infants aged 9 months had 0.089/0.18 µg/kg bw/week mean/P90 lower-bound exposure and 0.14/0.29 µg/kg bw/week mean/P90 upper-bound exposure (p. 59).
FSANZ reported that methylmercury exposure for women of childbearing age and all assessed groups except children aged 2-5 years did not exceed the PTWI; children aged 2-5 years reached up to 110% of PTWI at the mean and 220% of PTWI at P90 (p. 60).

Methods (brief)

FSANZ sampled 88 foods and beverages, including tap water, across Australia and analysed 508 composite samples. Foods were purchased in all states and territories during May 2013 and February 2014, prepared to a ready-to-eat state, and composited after preparation (pp. 16-18). All foods were analysed for total arsenic, cadmium, lead, and total mercury. Inorganic arsenic was analysed only in takeaway fish fillets, frozen fish portions, mussels, prawns, rice, sushi, and canned tuna in brine; inorganic and methylmercury were analysed only in takeaway fish fillets, frozen fish portions, mussels, prawns, and canned tuna in brine (Table 1, p. 17).

Analytical methods were NATA-accredited for agricultural chemicals and total metal contaminants, except for inorganic arsenic and organic mercury methods, which were validated and pending accreditation at the time of 2013-2014 testing; Symbio Laboratories later obtained NATA accreditation for those tests. Cadmium, lead, mercury, and total arsenic used ICP-MS after acid digestion with LOR 0.005 mg/kg (drinking water LOR 0.0001 mg/L). Inorganic arsenic used HPLC-ICP-MS after freeze-dried enzymatic digestion with LOR 0.01 mg/kg. Inorganic and methylmercury used HPLC/ICP-MS after freeze-dried preparation with LORs of 0.01 mg/kg for inorganic mercury and 0.05 mg/kg for methylmercury (Table 2, pp. 18-19).

For dietary modelling, FSANZ combined concentration data with food-consumption data from Australian nutrition surveys. Metal-contaminant exposure modelling used median concentrations for each food; food-concentration comparison tables report means with non-detects assigned to the LOR. Dietary exposure ranges use lower-bound (nd=0) and upper-bound (nd=LOR) scenarios, not a single half-LOR substitution (pp. 20-23, 34, 42, 48, 54, 59).

Limitations

The source is a total-diet study, not a commodity-specific market survey. It is strong for broad dietary exposure and Australian regulatory context but less granular for species, brand, region, or supplier-specific risk within seafood categories. Speciation coverage was limited to targeted foods: inorganic arsenic was not measured across the whole food basket, and inorganic/methylmercury were measured only in selected seafoods. Sampling occurred in 2013-2014; the report was published in June 2019, so current occurrence patterns may differ.

Several risk-characterisation statements are FSANZ’s own source-side conclusions. This wiki page preserves those statements as findings of the source and does not convert them into consumer advice or HMT&C threshold claims.

Implications

Certification: Provides source-side Australian/New Zealand regulatory and exposure context for arsenic, cadmium, lead, and mercury in broad foods, with direct seafood relevance for tAs, tHg, iHg, and MeHg. It contributes occurrence and dietary-exposure evidence; it does not propose certification thresholds.
Courses: Useful teaching example for total-diet-study design, non-detect scenario handling, and speciation discipline across tAs/iAs and tHg/iHg/MeHg.
App: Supports broad exposure context for seafood-driven total arsenic and methylmercury, rice-driven inorganic arsenic, and root-vegetable/grain-driven cadmium. It does not provide an ingredient-level distribution for every food in the basket.
Microbiome: Not a primary microbiome source.

Wiki pages this source may touch

Verification notes

Codex merge-enhanced this page on 2026-05-18 from manual-fetch PDF P0160. The P0160 PDF SHA-256 (529f57593c5918a3be9112c061dea3e084455531e7ac464e179f4d0742c6391f) is byte-identical to the earlier raw report path raw/reports/fsanz2019-25th-australian-total-diet-study.pdf; this is a provenance merge, not a true duplicate conflict.
Legacy wording said “half the LOR used” for all non-detects. Source methods instead use lower-bound nd=0 and upper-bound nd=LOR scenarios for metal dietary exposure, while international comparison tables may use source-specific middle-bound values for comparator datasets. Methods and Key numbers were corrected accordingly.
The cadmium dietary-exposure paragraph on p. 42 reports the upper-bound P90 range as “8.8-20 µg/kg bw/day” while the section, PTMI comparator, and adjacent values are monthly. This page records the printed range and flags the unit conflict rather than silently normalizing it.
The source’s “acceptably low,” “not of concern,” and fish-consumption-benefit statements are source-side FSANZ risk characterisation. This source page frames them as FSANZ findings and avoids consumer advice or HMT&C threshold recommendations.
No sampled product brands are named in this page. Scientific-method laboratory/vendor names are retained only where needed for reproducibility and fall under the Part 12 methods exception.
Fresh-context audit (Codex, 2026-05-18) found the page numerically clean and firewall-clean, but flagged iHg as missing from the GPT metal vocabulary. The source page keeps iHg because FSANZ measured inorganic mercury separately (pp. 56-57); the vocabulary was updated to permit iHg.

Page history

The five most recent substantive edits to this page. The full version history lives in git; when DOI minting comes online (see schema docs), each entry below will also link to a version-pinned DataCite DOI.

Commit	Date	Description
b0f3d38	2026-06-12	batch \| corpus rescreen b04 old terminal skips

Heavy Metal Index

Index