FDA 2016 — Arsenic in rice and rice products risk assessment report
FDA’s Center for Food Safety and Applied Nutrition (CFSAN) released this risk assessment in March 2016 (revised version for public comment; earlier drafts from 2013–2015). The report characterizes lung and bladder cancer risk from inorganic arsenic in rice and rice products for the U.S. population, using FDA’s own dose-response model and exposure assessment data from FDA’s 2013 and 2015 arsenic-in-rice surveys, Consumer Reports survey data, and dietary consumption data. The report was the key scientific basis for FDA’s proposed action level for inorganic arsenic in infant rice cereal (100 ppb iAs) that was finalized in 2020 under the Closer to Zero program. The concentration data section includes average iAs by product category for infant/toddler foods, which is among the most comprehensive FDA surveillance datasets for rice-based infant food.
Key numbers
Average inorganic arsenic in infant/toddler foods (from FDA survey data):
| Product Category | Average iAs (ppb) | Range iAs (ppb) | n |
|---|---|---|---|
| Cereal – Infant/Toddler (rice only) | 103 | 20.8–176 | 76 |
| Cereal – Infant/Toddler (multigrain) | 30.0 | 22.2–49.6 | 6 |
| Cereal – Infant/Toddler (non-rice) | 13.9 | 3.5–68.3 | 30 |
| Apples | 3.0 | 3.0–3.0 | 10 |
| Cereal – Oat Ring | 17.2 | 4.4–59.6 | 30 |
| Grapes | 4.5 | 3.0–15.2 | 10 |
| Juice – Apricot | 1.1 | 0.05–3.0 | 29 |
| Juice – Grape | 12.4 | 0.6–49.6 | 61 |
| Juice – Peach | 2.6 | 0.3–15.2 | 29 |
| Juice – Pear | 4.3 | 1.0–9.7 | 30 |
| Juice Boxes and Pouches | 5.5 | 0.9–32.1 | 40 |
| Peanut Butter | 3.4 | 3.0–5.9 | 29 |
| Quinoa | 7.9 | 0.3–49.0 | 30 |
| Raisins | 8.1 | 3.0–22.6 | 23 |
| Stage 2 Toddler Foods | 8.1 | 0.3–21.3 | 35 |
| Teething Biscuits | 44.4 | 17.0–71.5 | 27 |
| Toddler Puffs | 53.9 | 16.4–145 | 31 |
Key finding: Rice-based infant cereal average iAs 103 ppb (n=76, range 20.8–176 ppb) is approximately 7× higher than non-rice infant cereal (13.9 ppb) and 3.4× higher than multigrain cereal (30.0 ppb). Toddler puffs (53.9 ppb) and teething biscuits (44.4 ppb) are second and third highest, both rice-containing.
Full sample-level data table in appendix covers 76 rice cereal samples with individual iAs values (20.8–176 ppb); brown rice cereal samples cluster at the upper end (typical 87–176 ppb range); white rice cereal samples lower (typical 75–112 ppb range).
Methods (brief)
Multi-year risk assessment integrating: (1) FDA 2013 and 2015 arsenic-in-rice surveys providing iAs concentration data; (2) Consumer Reports survey data; (3) U.S. dietary consumption data from NHANES. Hazard characterization: FDA dose-response model for lung and bladder cancer from inorganic arsenic based on Taiwanese epidemiological data; compared against U.S. EPA drinking-water dose-response model and Liao et al. (2009) model. Exposure assessment: iAs concentration × food consumption amounts from NHANES. Below-detection-limit handling: half the detection limit substituted for total As samples below detection; iAs values below detection set to half the detection limit. Analytical method: ICP-MS with HPLC speciation for iAs. Report underwent extensive peer review (interagency, external).
Limitations
This is a version released for public comment (March 2016); it is not the final FDA risk assessment. The risk assessment focuses on lung and bladder cancer endpoints and does not quantify non-cancer health effects (developmental neurotoxicity, cardiovascular effects). Country-of-origin labeling for rice cereal was “Not Specified” for most samples; geographic source breakdown of the 76 rice cereal samples is unavailable. The dose-response model is based on Taiwanese population data from high drinking-water As exposure, with uncertainty in extrapolation to lower dietary exposure from rice. The FDA’s subsequently finalized action level of 100 ppb for infant rice cereal is consistent with the average concentration in this dataset (103 ppb mean).
Implications
- Certification: This is the authoritative source for FDA’s iAs surveillance data on infant rice cereal and the scientific basis for the 100 ppb CTZ action level. Any HMT&C infant-cereal standard must reference this dataset. The 103 ppb average in 76 rice cereal samples, with a range of 20.8–176 ppb, establishes the market baseline against which CTZ’s 100 ppb limit was set.
- Courses: Essential regulatory reference — demonstrates how FDA used both average and distribution data to set an action level; illustrates the difference between rice and non-rice cereal iAs levels in the infant food category.
- App: Note only — contamination_profile for rice (infant cereal matrix). Average iAs 103 ppb, range 20.8–176 ppb for rice-only infant cereal (n=76). Non-rice infant cereal average 13.9 ppb for comparison.
- Microbiome: Not applicable.