FSAI 2016 — Total Diet Study Ireland 2012–2014

This is the second Total Diet Study (TDS) conducted by the Food Safety Authority of Ireland, covering chemical exposure in the Irish population from food purchased in autumn 2012. 141 food samples (1,043 sub-samples) were collected from major Dublin retailers, prepared as consumed, and analysed under contract by the Food and Environment Research Agency (FERA, UK). Dietary exposure was modelled by combining measured occurrence levels with food consumption data from the National Adult Nutrition Survey (NANS, n = 1,500 adults aged 18+, 2008–2010) and the National Children’s Food Survey (NCFS, n = 594 children aged 5–12, 2003–2004). The report covers contaminant metals (Al, total As, inorganic As, Cd, Cr, Pb, total Hg, Sn), essential nutrients (I, Se), food additives (nitrates, nitrites), acrylamide, mycotoxins, PAHs, BPA, phthalates, and pesticides. For heavy metals, Irish dietary exposures were below health-based guidance values for all metals except cadmium, where exceedances of the EFSA TWI were observed at the 97.5th percentile in both adults and children. Cereals were the dominant contributor to inorganic-arsenic, lead, and aluminium exposure; fish (white and canned) dominated mercury exposure.

Key numbers

All exposure values per the report’s headline tables (Tables 3, 4, 5, 8, 9, 11, 13, 15); ranges given as LB–UB (lower-bound, treating < LOD as zero, to upper-bound, treating < LOD as LOD). All comparisons to health-based guidance values use the Table values directly.

Aluminium (Table 3): Adult mean intake 0.05 mg/kg bw/day (0.35 mg/kg bw/week), 35% of EFSA TWI (1 mg/kg bw/week); 97.5th percentile 0.12 mg/kg bw/day (0.83–0.84 mg/kg bw/week), 83–84% of TWI. Child mean intake 0.05 mg/kg bw/day (0.36–0.37 mg/kg bw/week), 36–37% of TWI; 97.5th percentile 0.11 mg/kg bw/day (0.74–0.75 mg/kg bw/week), 74–75% of TWI. Aluminium detected in 84% of samples. Major contributors (LB): non-alcoholic beverages 40% and cereals 33% for adults; cereals 54% and non-alcoholic beverages 10% for children.

Total arsenic (Table 4): Adult mean 0.7–0.9 µg/kg bw/day; 97.5th percentile 3.9–4.2 µg/kg bw/day. Child mean 0.6–0.9 µg/kg bw/day; 97.5th percentile 2.9–3.3 µg/kg bw/day. Total arsenic detected in 17% of samples. Fish and seafood contributed 95% (adults) and 89% (children) of total arsenic intake — predominantly as the low-toxicity organic forms (arsenobetaine, arsenosugars).

Inorganic arsenic (Table 5): Adult mean 0.01–0.02 µg/kg bw/day; 97.5th percentile 0.06–0.08 µg/kg bw/day. Child mean 0.03–0.05 µg/kg bw/day; 97.5th percentile 0.13–0.14 µg/kg bw/day. Of the foodstuffs that contained detectable total As, 33% (eight samples) contained iAs above LOD. Exposures fall below the EFSA BMDL01 range of 0.3–8 µg/kg bw/day for cancers of the lung, skin, and bladder, but margins of exposure (Table 7) are low: vs. BMDL01 0.3 µg/kg bw/day MoEs range from 2 (children 97.5th percentile) to 30 (adult mean); vs. BMDL01 8 µg/kg bw/day MoEs range from 57 (children 97.5th percentile) to 800 (adult mean). Major contributors to iAs: cereals 81% (adults) and 94% (children); fish and seafood contributed only 4% (adults) and 0.3% (children) of iAs despite dominating total As.

Cadmium (Table 8): Adult mean 0.16–0.22 µg/kg bw/day (1.1–1.5 µg/kg bw/week), 44–62% of EFSA TWI (2.5 µg/kg bw/week); 97.5th percentile 0.33–0.42 µg/kg bw/day (2.3–3.0 µg/kg bw/week), 92–118% of TWI. Child mean 0.24–0.32 µg/kg bw/day (1.7–2.3 µg/kg bw/week), 66–91% of TWI; 97.5th percentile 0.47–0.59 µg/kg bw/day (3.3–4.1 µg/kg bw/week), 132–164% of TWI. Cadmium detected in 43% of samples. Major contributors: cereals 39% and vegetables 36% (adults); cereals 48% and vegetables 30% (children). Authors note this is a considerable reduction vs. the previous Irish TDS (which estimated adult mean 95–123% TWI and 97.5th percentile 216–244% TWI), attributed to a roughly 50% drop in potato consumption between the IUNA (2001) and NANS (2008–2010) adult food consumption surveys. A urinary biomarker study from NANS participants found 95% of the population (including women aged >50, the most at-risk subgroup) below the EFSA critical value of 1 µg Cd/g creatinine, supporting the FSAI conclusion that long-term chronic Cd exposure does not present an unacceptable risk.

Chromium (Table 9, total Cr; no Cr-VI speciation): Adult mean 0.7–1.2 µg/kg bw/day, 0.2–0.4% of EFSA TDI (300 µg Cr(III)/kg bw/day); 97.5th percentile 1.4–2.1 µg/kg bw/day, 0.5–0.7% of TDI. Child mean 1.2–1.8 µg/kg bw/day, 0.4–0.6% of TDI; 97.5th percentile 2.5–3.2 µg/kg bw/day, 0.6–1.1% of TDI. Chromium detected in 48% of samples. Major contributors: vegetables 31%, meat 26%, cereals 10% (adults); vegetables 26%, cereals 17%, meat 16%, non-alcoholic beverages 16% (children). Risk assessment assumes all dietary chromium is Cr(III); Cr-VI in food was not measured (EFSA assessed Cr-VI exposure only from drinking water).

Lead (Table 11): Adult mean 0.04–0.12 µg/kg bw/day; 97.5th percentile 0.11–0.22 µg/kg bw/day. Child mean 0.04–0.17 µg/kg bw/day; 97.5th percentile 0.09–0.27 µg/kg bw/day. Lead detected above LOD in 29% of samples. Margins of exposure (Table 12): adults 5–18 vs. BMDL10 0.63 µg/kg bw/day for chronic kidney disease; 3–6 for 97.5th-percentile; 13–38 vs. BMDL01 1.50 µg/kg bw/day for systolic blood pressure; 7–14 for 97.5th-percentile. Children 3–13 (mean) and 2–6 (97.5th-percentile) vs. BMDL01 0.50 µg/kg bw/day for developmental neurotoxicity. Major contributors: alcoholic beverages 28%, cereals 22%, vegetables 12% (adults); cereals 37%, non-alcoholic beverages 19%, vegetables 12% (children). Tap water samples from individual households did not contain detectable lead; however, the report flags legacy lead piping in up to 200,000 Irish properties (Irish Water, 2015b) as a separate exposure pathway. The 1998 Drinking Water Directive sets a 10 µg/L limit for lead in drinking water (98.7% compliance reported in Ireland for 2014).

Mercury (Table 13, total Hg; no MeHg speciation): Adult mean 0.02–0.12 µg/kg bw/day (0.17–0.84 µg/kg bw/week); LB 10% of EFSA MeHg PTWI (1.6 µg/kg bw/week), UB 52% (asterisked: UB includes < LOD assumed at LOD across all food groups including non-fish, so overstates MeHg). 97.5th percentile 0.12–0.25 µg/kg bw/day (0.84–1.74 µg/kg bw/week), 53–109% (UB asterisked). Child mean 0.02–0.17 µg/kg bw/day (0.14–1.19 µg/kg bw/week), 9–74% (UB asterisked). Mercury was found in 11% of samples and 97% of those were fish and seafood, so the LB estimate is the most valid comparison. On LB basis, adult mean MeHg exposure 10% of PTWI and 97.5th percentile 53%; child mean 9% and 97.5th percentile 48%. Major contributors: white fish 52% and canned fish 29% (adults); white fish 59% and canned fish 36% (children); fish and seafood together accounted for 96% (adults) and 99% (children) of total Hg intake. The FSAI provides standing consumption advice for children, pregnant women, and women of reproductive age regarding predatory fish (shark, marlin, swordfish, fresh tuna). A pilot biomonitoring study (DEMOCOPHES, Cullen et al. 2014, n = 120 Irish mother/child pairs) reported mean hair Hg of 0.262 µg/g (mothers) and 0.149 µg/g (children), both below the US EPA guidance value of 1.0 µg/g.

Tin (Table 15): Adult mean 0.005–0.01 mg/kg bw/day (0.03–0.04 mg/kg bw/week), 0.24–0.29% of JECFA PTWI (14 mg/kg bw/week); 97.5th percentile 0.04 mg/kg bw/day (0.28–0.29 mg/kg bw/week), 2.03–2.07% of PTWI. Child mean 0.01 mg/kg bw/day (0.05–0.06 mg/kg bw/week), 0.38–0.5% of PTWI; 97.5th percentile 0.07 mg/kg bw/day (0.48–0.49 mg/kg bw/week), 3.4–3.5% of PTWI. Tin detected in only 9% of foods, of which 73% were canned. Major contributors: canned fruit 49%, canned vegetables 27%, canned soups 24% (adults); canned fruit 49%, canned soups 28%, canned vegetables 23% (children).

Per-category occurrence ranges (Annex III, Table 25; LB–UB, mg/kg fresh weight): Selected categories of HMI interest:

• Fats & Oils (olive oil, vegetable oil, fat, hard cooking fat, non-dairy spreads): Al 0–0.20, Cr 0–0.06, total As 0–0.02, Se 0–0.02, Cd 0–0.01, Sn 0–0.10, Hg 0–0.01, Pb 0–0.01 mg/kg. iAs was not analysed for Fats & Oils (Annex II scope for olive oil and vegetable oil excludes iAs).
• Cereals: Al 0.60–47.1, Cr 0–0.15, total As 0–0.03, iAs 0–0.02, Cd 0–0.14, Sn 0–0.10, Hg 0–0.01, Pb 0–0.01 mg/kg.
• Fish and Fishery Products: Al 0.10–71.7, total As 0.35–4.1, iAs 0–0.05, Cd 0–0.10, Sn 0–0.05, Hg 0.01–0.11, Pb 0–0.22 mg/kg.
• Fish and Fishery Products (canned): Al 0.70–8.1, total As 0.39–2.0, Cd 0–0.02, Sn 0–0.07, Hg 0.02–0.15, Pb 0–0.02 mg/kg.
• Vegetables: Al 0–4.0, Cr 0–0.15, total As 0–0.02, iAs 0–0.01, Cd 0–0.07, Pb 0–0.01 mg/kg.
• Fruit (canned): Sn 78.8–124 mg/kg (the dominant tin signal).
• Tea & Coffee: Al 0.05–3.4 mg/kg.

Selenium and iodine: Adult mean iodine intake 147–153 µg/day, 24.5–25.4% of SCF UL (600 µg/day); child mean 162–167 µg/day, 27.1–27.8% UL. Dairy contributes 73% (adults) and 85% (children) of iodine intake. Adult mean selenium intake very close to EFSA AI of 70 µg/day; child mean 34.7–39 µg/day. Neither deficiency nor excess concern for the Irish population.

Methods

TDS design following the harmonised WHO/EFSA/FAO TDS guidance (EFSA/WHO/FAO, 2011). Foods purchased in Dublin in autumn 2012; tap water sampled from a variety of private households connected to the public water supply. Sub-samples (typically five per food) selected on brand-market-share information from the food consumption databases; foods prepared as consumed (cooked, grilled, baked) before analysis; the Creme Food semi-probabilistic exposure-modelling software combined single aggregate-sample occurrence data with NANS/NCFS food intake distributions. Per-kg-bw calculations weighted by individual survey-participant body weights. Lower-bound (LB; < LOD = 0) and upper-bound (UB; < LOD = LOD) treatments of left-censored data applied throughout.

Analytical work by FERA (UK); methods accredited. For the heavy-metals analytes, ICP-MS was used for all eight metals (Annex I): typical LOD/LOQ Al 100/333, Pb 5/17, Cd 5/17, total Hg 5/17, total As 10/33, iAs 10/33, Sn 50/167, Cr 30/100, Se 10/17 µg/kg (ppb).

Speciation: Arsenic was speciated into total As and inorganic As where iAs was within scope (cereals and cereal products, rice, fish, herbs, spices, snacks containing iAs, several condiment categories; Annex II). Fats & Oils were analysed for total As but not iAs. Mercury was reported as total Hg only; no methylmercury speciation, although the report’s risk characterisation conservatively assumes that all mercury detected in fish is MeHg. Chromium was reported as total Cr; Cr-VI speciation not performed (the report assumes all Cr in food is the less toxic Cr(III), with Cr-VI exposure assessed separately by EFSA only from drinking water).

Implications

Certification: Provides the Irish national heavy-metal exposure baseline for Al, tAs, iAs, Cd, Cr, Pb, tHg, and Sn across the full retail diet, with food-category occurrence data (Annex III Table 25) in native fresh-weight basis. The Cd 97.5th-percentile exceedance of the EFSA TWI (92–118% adults, 132–164% children) is the only finding that crosses a health-based guidance value, and is driven mainly by cereals (39–48% contribution) and vegetables (30–36%). For the oils category specifically, all eight heavy metals are present at the lower end of their fresh-weight distributions and iAs was not analysed, limiting this report’s direct utility for oils threshold-setting on iAs.

Courses: Useful teaching case for the TDS methodology — how a representative national food consumption survey, food sampling, preparation-as-consumed, occurrence measurement, and probabilistic exposure modelling combine to produce population-level estimates against health-based guidance values. Also illustrates the iAs/tAs speciation imperative: total arsenic in this study is dominated by fish (95% adults), but inorganic arsenic is dominated by cereals (81–94%) because fish arsenic is overwhelmingly organic. Conflating the two would misallocate exposure attribution by roughly an order of magnitude.

App: Irish dietary exposure data for adults and children across a broad food list (Annex II), with per-food-group LB–UB occurrence ranges (Annex III, Table 25) suitable for ingredient- and category-level reference values.

Verification notes

2026-05-26: Page enhanced from earlier ingest. Defects corrected:

• raw_path updated to the actual PDF filename (previous value was truncated to …Authority of Ireland (FSAI TDS).pdf → …Authority of Ireland (FSAI TDS).pdf full filename including the full Food Safety Authority of Ireland title).
• access_url added (FSAI publication URL).
• ingredients cleanup: removed tuna (no wiki/ingredients/tuna.md exists). Added canned-tuna and tinned-fish, the two valid ingredient slugs for the fish-and-fishery-products subcategories actually sampled by the TDS (Annex II: “Canned Tuna” and “Tinned Fish excl. salmon & tuna” are distinct sub-samples) and dominant contributors to mercury exposure on the report’s own breakdown.
• products cleanup: removed infant-cereal. FSAI TDS Annex II does not include infant cereal as a sampled food category (the cereal scope is white flour, wholemeal flour, breads, biscuits, cakes, pasta, rice, breakfast cereals — no infant cereal). canned-fish retained as a valid product-category routing target.
• Chromium contributor attribution for children corrected: prior page said “fruit juices (16%)”, but Figure 4 (children) shows non-alcoholic beverages 16%; fruit juices are a subset but not the whole. Updated to “non-alcoholic beverages 16%“.
• Methods section enhanced with ICP-MS attribution and per-analyte LOD/LOQ values from Annex I.
• Per-food-category occurrence ranges added for Fats & Oils, Cereals, Fish, Canned Fish, Vegetables, Canned Fruit, and Tea & Coffee from Annex III Table 25, including the explicit note that iAs was not analysed for Fats & Oils.
• Tin contributor breakdown added (canned fruit 49%, canned vegetables 27%/23%, canned soups 24%/28%).
• Cadmium contributor breakdown added (cereals 39%/48%, vegetables 36%/30%).
• Mercury Table 13 figures updated to include the asterisked UB caveat from the source: UB values include non-fish food groups (where MeHg was not detected) and so overstate MeHg exposure; the LB estimate is the appropriate comparison to the MeHg PTWI per the report’s own framing.
• Lead margin-of-exposure values added per Table 12; iAs margin-of-exposure values added per Table 7.
• Implications → Certification: removed the phrase “consistent with European TDS literature” (Part 2 cross-source synthesis comparison). Reframed the oils-specific limitation (iAs not analysed) so the page is honest about what this source does and does not support for the oils category.
• updated set to 2026-05-26.

2026-05-26 audit pass: Auto-audit subagent verdict REVISE. Findings applied:

• Check 1 ❌ iAs MoE attribution corrected. The earlier wording (“2–62 across population subgroups vs. BMDL01 8 µg/kg bw/day; 2–30 vs. BMDL01 0.3 µg/kg bw/day”) conflated two different denominator columns of Table 7 (p.28). Per Table 7 the cross-subgroup range vs. BMDL01 8 is 57–800 (children 97.5th percentile UB → adult mean LB), and the cross-subgroup range vs. BMDL01 0.3 is 2–30. Page now reports both ranges with endpoint attribution.
• Check 2 ❌ Three dead-wikilink slugs corrected: ingredients/dairy → ingredients/milk-and-dairy (the canonical page; dairy is an alias the routing audit resolved via the alias index, but the rendered wikilink would not navigate); metals/aluminium → metals/aluminum (canonical filename uses US spelling); regulations/eu-1881-2006-contaminants → regulations/eu-1881-2006-contaminants-superseded (the page on disk).

Findings rejected (false positives):

• Check 3 ⚠️ FERA attribution flagged as “cannot verify from extracted pages.” Verified against source p.12: “Planning and co-ordination of this project as well as sampling of the foods of interest was undertaken by FSAI staff; food preparation and analysis of the samples was undertaken under contract by the Food and Environment Research Agency (FERA) in the UK.” FERA attribution kept.
• Check 5 ⚠️ “Conflating the two would misallocate exposure attribution by roughly an order of magnitude” flagged as synthesis-adjacent. The contrast is between two figures inside this single source (tAs fish dominance 95% vs. iAs cereal dominance 81–94%) — i.e., the source’s own data, not cross-source synthesis. Sentence kept.
• Check 2 audit also flagged ingredients/meat and ingredients/dairy against a possibly-stale taxonomy snapshot; ingredients/meat.md exists on disk so meat is kept; the dairy correction is for the dead-link reason above, not snapshot mismatch.

Wiki pages updated on ingest

• lead
• cadmium
• mercury
• aluminum
• arsenic
• chromium
• tin
• rice
• cereals
• fish
• canned-tuna
• tinned-fish
• olive-oil
• vegetable-oil
• canned-fish
• eu-1881-2006-contaminants-superseded

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.