Arsenic in US chicken from feed drugs

Overview

Nachman KE et al.. Environ Health Perspect 2013 · PMID 23694900 · DOI 10.1289/ehp.1206245

Conventional chicken had about three times more inorganic arsenic than organic chicken; estimated extra cancer risk was small.

Does the study support the specific claim? Yes · Methodological quality: moderate

Original publication

Plain-language summary

Scientific in-depth review

1. Full reference

Authors

Nachman KE, Baron PA, Raber G, Francesconi KA, Navas-Acien A, Love DC

Journal

Environ Health Perspect

Year

2013

Identifier

PMID 23694900 DOI 10.1289/ehp.1206245 · PMC3701911

Verification

verified – PubMed E-utilities efetch, 2026-10-25

Retracted

no

Corrections

PubMed lists: ErratumIn (Environ Health Perspect. 2013 Jul;121(7):823)

Extent checked

full text

Conflicts of interest and funding

Directed research grant from the Johns Hopkins Center for a Livable Future (P.A.B. fellowship from the same centre); authors declare no actual or potential competing financial interests. Published correction of unit and LADD errors (stated not to affect results).

2. Research question

What is the arsenic species profile in US chicken and the cancer risk from arsenic-based drugs?

3. Study design and population

Study design

Measurement / residue study – Market-basket sampling of chicken meat with arsenic speciation and cancer risk estimation

Population

Conventional, antibiotic-free and organic retail chicken in 10 US metro areas

Country

USA

Period

December 2010-June 2011

Sample size

142 chicken breast samples analysed for total arsenic (69 conventional, 34 conventional antibiotic-free, 37 USDA Organic; 60 brands, 82 stores); speciation in 78 cooked and 65 raw samples with total As >=10 ug/kg dry weight (cooked: 40 conventional, 13 antibiotic-free, 25 organic)

Inclusion and exclusion criteria

Chicken breast packages purchased from 5-10 grocery stores in each of 10 US metropolitan areas (max one package per brand per store) before the July 2011 roxarsone sales suspension; speciation restricted for budget reasons to samples with total As >=10 ug/kg DW. Only breast meat analysed.

4. Methods and endpoints

Exposure / intervention

Production type by label (conventional, antibiotic-free, organic); producer arsenical-drug policy (no known policy vs prohibitory policy, determined by contacting companies); roxarsone detection

Comparator

Organic and antibiotic-free chicken

Endpoints

iAs, roxarsone; estimated cancer cases

Statistical methods

Geometric means with 95% CIs; paired t-tests on log-transformed raw vs cooked pairs; Spearman correlations; deterministic cancer risk estimate = LADD (82.4 g/day chicken, 80 kg BW) x proposed (draft, not finalised) EPA IRIS slope factor 25.7 per mg/kg/day; values <LOD imputed as LOD/sqrt(2); lab blinded to cooked/raw status; 10 blinded duplicates.

5. Actual results

Results

Cooked chicken: total As GM 3.0 ug/kg (2.5-3.6). iAs GM conventional 1.8 (1.4-2.3) vs antibiotic-free 0.7 (0.5-1.0) vs organic 0.6 (0.5-0.8) ug/kg; producers with no known arsenical policy 2.0 (1.6-2.5) vs prohibitory policy 0.7 (0.5-0.9), p=0.0004. Roxarsone detected in 20 of 40 conventional (19 of 40 cooked in the results text), 1 of 13 antibiotic-free, 0 of 25 organic; iAs roxarsone-positive 2.3 (1.7-3.1) vs negative 0.8 (0.7-1.0). Cooking raised iAs by 27.2% (raw 0.8 to cooked 1.1, p=0.004) and lowered roxarsone by 44.1%. iAs was <1 ug/kg in 94% of organic and 93% of conventional samples from producers with prohibitory policies, but >1 ug/kg in 70% of samples from producers without such policies.

Absolute risks

Modelled: conventional chicken from producers without arsenical-prohibiting policies vs organic adds 0.115 ug iAs/day, i.e. about 3.7 additional lifetime bladder and/or lung cancers per 100,000 persons with lifetime exposure; extrapolated to 8,661 cases over 70 years (~124/year) if arsenicals were used in all US poultry. Based on a proposed, not finalised, slope factor derived mainly from women.

Authors' conclusion

Conventional chicken had higher iAs; cessation of arsenical drugs could reduce exposure.

6. Strengths

• Arsenic speciation by HPLC-ICP-MS with validated QC, blinded lab analysis and duplicate measurements
• Paired raw/cooked samples showing cooking effects
• 10 US cities, 60 brands; producer policy information gathered
• First quantification of roxarsone residues in chicken meat

7. Weaknesses

• Speciation only in samples with total As >=10 ug/kg DW (selection of higher-arsenic samples)
• Breast meat only; one sampling period (2010-2011) before US withdrawal of roxarsone
• Cancer estimate relies on a draft slope factor, average intake, lifetime exposure and the difference vs organic; no human biomarker or outcome data
• Small subgroup sizes (e.g. 13 antibiotic-free speciated)
• Minor internal inconsistency (roxarsone 19 vs 20 of 40 conventional samples)
• Funder is an academic centre with a stated food-system reform mission

8. Risk of bias

Orienting tool

none (exposure measurement and risk assessment; no standard RoB tool applies)

Type of assessment

structured assessment based on the full text by a single reviewer (no formal double review)

Full formal assessment

no

Overall risk of bias

some concerns

Rationale

Unchanged (some_concerns): measurements robust, risk extrapolation uncertain. Single-reviewer structured judgement; a formal assessment would need two independent reviewers.

Full-text review (25 October 2026): Added sampling frame (142 breasts, 60 brands, 82 stores), speciation threshold, producer-policy comparison (2.0 vs 0.7 ug/kg), cooking +27% iAs, 70% vs <7% exceeding 1 ug/kg, draft slope factor and population extrapolation (124 cancers/year), funding (JHU Center for a Livable Future), correction notice. Bias unchanged.

9. Assessment of the statistical interpretation

Relative differences in iAs (about 3-fold) are statistically robust, but absolute concentrations are very low (~1-2 ug/kg). The cancer figure is a model-based upper-bound-style estimate with considerable uncertainty acknowledged by the authors, not an observed effect.

10. Relevance to the specific claim

C-CHK-07.

11. Independent contrary findings

I-CHK-03 (withdrawal).

12. Transferability to Germany and Europe

Low: not authorised in EU.

13. Overall methodological judgement

Methodologically sound exposure study showing that roxarsone use in US chicken (2010-2011) increased low-level inorganic arsenic in meat; the cancer burden is modelled and uncertain. Historically relevant to the US; arsenicals are not authorised in the EU.

14. Open questions

Current use of arsenicals in poultry in other countries including Egypt; contribution of chicken to total iAs exposure compared with rice and water.

15. Linked fact checks

• Chickens were fed arsenical drugs (roxarsone) that raise carcinogenic inorganic arsenic in chicken meat and in consumers. – Studies supporting the claim

Use by the Tayebat archive

Note: according to the archive, the studies were compiled by the archive operators. Whether Dr. Elawady himself cited this study is not established.

What the archive says about the study

Conventional US chicken contained two to three times the inorganic arsenic of organic chicken; roxarsone-positive samples had higher carcinogenic arsenic; cooking raised concentrations.

A. What was actually studied?

Arsenic species in US chicken.

B. What did the authors find?

Cooked conventional chicken iAs GM 1.8 vs 0.6 ug/kg in organic (about 3-fold); roxarsone in 20/40 conventional, 0/25 organic; cooking increased iAs by 27%; modelled 3.7 extra lifetime cancers per 100,000 with lifetime exposure (draft slope factor).

C. What conclusions do the authors themselves draw?

Stopping arsenicals could reduce exposure.

D. What does the archive derive from it?

Used for "raised on arsenical drugs".

E. Is this derivation scientifically justified?

Essentially correct for the US 2010-2011 situation: numbers match. The archive omits that absolute iAs levels were ~1-2 ug/kg, that the cancer estimate is modelled with a draft slope factor, that the leading US roxarsone marketer suspended sales in July 2011 (after sampling), and that arsenicals are not authorised in the EU.

Adequacy of the interpretation (by the archive)

essentially correct but incomplete