Heavy Metal Index
Ask the IndexAboutIndex Pro

Children's Cadmium-Containing Metal Jewelry — Staff Report

Cite this pageVersion history

Overview

This 109-page U.S. Consumer Product Safety Commission staff document was transmitted on October 19, 2010 from Colin Church (Voluntary Standards Coordinator) to Brent Cleaveland (Chairman, ASTM F-15.24 Subcommittee on Children’s Jewelry), to support development of an ASTM voluntary safety standard for cadmium in children’s metal jewelry. The body of the document comprises three CPSC staff memoranda dated October 14, 2010: (i) the Staff Report on Children’s Cadmium-Containing Metal Jewelry authored by Kristina M. Hatlelid, Ph.D., M.P.H. (Toxicologist, Directorate for Health Sciences); (ii) TAB A: Staff Responses to Peer Review Comments on the draft staff report; and (iii) TAB B: Toxicity Review of Cadmium authored by Dominique J. Williams and Kristina M. Hatlelid. All three memoranda carry the standard CPSC staff caveat that the contents “are those of the CPSC staff and have not been reviewed or approved by, and may not necessarily reflect the views of, the Commission.” The Staff Report’s role in the CPSC Federal Hazardous Substances Act (FHSA) regulatory regime is to derive an acceptable daily intake level (ADI) for cadmium, translate that ADI into per-item extraction-test thresholds that distinguish products that may meet FHSA hazardous-substance criteria from those that do not, and to set out the staff’s assessment approach for cadmium-containing metal jewelry intended for children. No primary jewelry-content measurements are reported in this document; the cover letter references separate jewelry test-result enclosures (“Tabs A and B” in the October 19, 2010 transmittal letter) that are not contained in this 109-page PDF file.

Key numbers

Cover-letter framing (Colin Church to Brent Cleaveland, October 19, 2010)

  • Originating CPSC staff position (p. 1 of cover letter): “CPSC staff concluded that a test method for chemicals that migrate from small items that can be swallowed, especially metal items, should be based on measuring solubility in an acidic solution over 24 hours. This conclusion is based on the results of testing hundreds of jewelry and metal alloy samples, as well as on information about the length of time an ingested foreign object could be present in the digestive tract of a child.”
  • CPSC staff position on existing two-hour toy-safety extraction methods (p. 1 of cover letter): “CPSC staff believes that the test methods in the toy safety standards ASTM F-963 and EN-71-3, which specify measuring solubility of the chemicals after two hours, are inadequate for evaluating potential chemical exposures from products such as metal jewelry.”

Staff Report (Hatlelid, October 14, 2010) — derived ADIs and test thresholds

  • Acceptable Daily Intake (ADI), chronic oral cadmium exposure: 0.1 µg/kg/day (p. 5 of Staff Report). Derived as the cadmium intake level of 0.33 µg/kg/day (the ATSDR 2008 PK-modelled intake corresponding to the Suwazono et al. 2006 urinary cadmium BMDL of 0.5 µg cadmium per gram creatinine) divided by an uncertainty factor of 3 to account for sensitive subpopulations (reduced from the default factor of 10 because of the strength of the supporting human-population evidence). “This is the level of chronic exposure that should not be exceeded to avoid adverse health effects.”
  • Acceptable Daily Intake (ADI), acute oral cadmium exposure: 11 µg/kg/day (p. 7 of Staff Report). Derived as the NOAEL of 1.1 mg/kg/day from Borzelleca et al. (1989) — a 10-day cadmium-chloride drinking-water study in Sprague-Dawley rats — divided by an uncertainty factor of 100 (10× for use of animal data instead of human data; 10× for sensitive individuals in the human population).
  • Body-weight assumption for young children (ages 2–6 years): 18.2 kg (≈ 40 lb), from Ogden et al. 2004 (p. 8 of Staff Report).
  • Maximum allowable chronic exposure for a young child: 1.8 µg/day (p. 8 of Staff Report). Computed as 0.1 µg/kg/day × 18.2 kg = 1.82 µg/day, reported as “about 1.8 µg/day.”
  • Maximum allowable acute exposure for a young child: about 200 µg/day (p. 8 of Staff Report). Computed as 11 µg/kg/day × 18.2 kg = 200.2 µg/day.
  • Mouthing-time assumption (ages 24–36 months, all objects excluding pacifiers): 37 minutes/day average daily mouthing time, from CPSC (2001) mouthing observation study (p. 9 of Staff Report).
  • Saline extraction threshold for mouthable cadmium-containing jewelry: a 6-hour saline-extraction laboratory result that exceeds 18 µg total cadmium indicates the product may meet FHSA hazardous-substance criteria based on chronic toxicity (p. 9 of Staff Report). Derivation: (1.8 µg/day) / (37 min/day) × (6 hours × 60 min/hour) = 18 µg total cadmium extracted in the 6-hour saline test (Eq. 1, p. 9 of Staff Report).
  • Acid extraction threshold for swallowable cadmium-containing jewelry: a 24-hour acid-extraction laboratory result that exceeds 200 µg total cadmium indicates the product may meet FHSA hazardous-substance criteria based on acute toxicity (p. 10 of Staff Report). Derivation: the acute ADI of 200 µg/day is treated as the total exposure that can occur within the short time an ingested item remains in the GI tract.
  • Laboratory protocol: CPSC’s standard laboratory procedure for evaluating possible mouthing exposure is 6-hour saline extraction; for evaluating possible swallowing exposure of metal jewelry items the protocol is 24-hour dilute hydrochloric acid extraction (p. 9 of Staff Report). The 24-hour acid extraction duration is based on data showing that ingested items sometimes remain in the stomach for several days; CPSC staff data indicate that 24 hours is generally sufficient to identify products that could leach large amounts of chemicals.
  • Critical exposure level used to derive the chronic ADI: cadmium intake of 0.33 µg/kg/day (p. 4 of Staff Report), from ATSDR (2008) PK modelling of European-population cadmium intake corresponding to a urinary cadmium concentration of 0.5 µg cadmium per gram creatinine — the lower bound of the BMDL (5 % excess risk of pHC proteinuria) reported by Suwazono et al. (2006) in 820 Swedish women.

Existing exposure limits surveyed by CPSC staff (Staff Report, p. 3)

  • WHO/JECFA Provisional Tolerable Monthly Intake (PTMI) for cadmium: 25 µg/kg body weight/month (JECFA 2010), equivalent to about 0.8 µg/kg/day.
  • U.S. EPA oral reference dose (RfD) for cadmium: 1 µg/kg/day for food intake; 0.5 µg/kg/day for cadmium in drinking water (EPA 1994 IRIS).
  • U.S. EPA cadmium-in-drinking-water limits: 0.04 mg/L for exposures up to 10 days; 0.005 mg/L for lifetime exposures (EPA 2003).
  • U.S. FDA bottled-water cadmium limit: 0.005 mg/L (FDA 2009).
  • OSHA occupational inhalation limit (8-hour TWA): 5 µg/m³ cadmium in air (OSHA 2009).
  • ATSDR chronic-duration oral Minimal Risk Level (MRL): 0.1 µg/kg/day (ATSDR 2008 draft toxicological profile), based on kidney effects in humans.
  • ATSDR intermediate-duration oral MRL: 0.5 µg/kg/day (ATSDR 2008 draft toxicological profile), based on effects on bone in experimental animals; no acute-duration MRL has been derived due to inadequacies in the acute oral exposure database.
  • European Standard EN 71-3 (1994) safety-of-toys cadmium migration limit: 0.6 µg/day cadmium from toys for young children, based on background cadmium intake levels and a policy decision to limit cadmium exposure from toys to 5 percent of the assumed background exposure.
  • European Council Directive 2009/48/EC (effective July 20, 2013) cadmium toy-safety limit: about 0.2 µg/day cadmium from toys (Staff Report notes these toy-safety daily limits are not body-weight-adjusted but are based on the weight of a young child, 7.5 kg ≈ 16.5 lb, for the revised standard).

Acute-toxicity literature reviewed (Staff Report, p. 5–6)

  • Documented severe-poisoning case (Nordberg et al. 1973): boys ages 13–15 years drank a beverage from a soda machine containing 16 mg/L cadmium. Onset of symptoms (headache, nausea, vomiting, abdominal pain, diarrhoea) within 15 minutes of consumption; symptoms resolved within 7 hours. Reported consumption: 0.5–2.5 glasses (average about one glass; ≈ 0.15 L), giving an estimated cadmium dose of about 2.4 mg total. CPSC staff considered the documentation inadequate for quantitative derivation of an exposure limit.
  • Acute oral LD50 for cadmium in rats and mice: approximately 100–300 mg/kg (Baer and Benson 1987; Kostial et al. 1978; Kostial et al. 1979).
  • Lowest dose found to cause death in rats: 15.3 mg/kg cadmium chloride administered as a single gavage dose in Sprague-Dawley rats (2 of 20 animals; Borzelleca et al. 1989).
  • Other acute oral effects in rats (Baranski 1985; Borzelleca et al. 1989; Machemer and Lorke 1981): reduced body weight or reduced body weight gain at cadmium doses greater than about 2 mg/kg/day; haemorrhages, ulcers, and reddening of the stomach and intestinal tract at 61 mg/kg/day; testicular atrophy at 66 mg/kg/day; necrotic changes in the kidney at 15.3 mg/kg/day and in the liver at 138 mg/kg/day.
  • Selected key acute study (Borzelleca et al. 1989): 10-day cadmium-chloride drinking-water exposure in male and female rats (strain not specified for the drinking-water arm in the Staff Report; the companion single-dose gavage arm in Borzelleca et al. 1989 used Sprague-Dawley rats, per TAB B p. 23); doses 1.1, 7.8, and 11.1 mg/kg/day in males and 1.1, 8.1, and 13.8 mg/kg/day in females. Dose-dependent decrease in body weight gain in males; reduced body weight in the highest-dose group in males; dose-related reduced body weight gain in females (no statistical analysis presented). A companion gavage study by the same authors at comparable doses also produced reduced body weight and reduced body weight gain, supporting that the effects are not attributable to drinking-water taste aversion. NOAEL = 1.1 mg/kg/day (the lowest dose tested in the drinking-water study), the basis for the 11 µg/kg/day acute ADI after a 100-fold uncertainty factor.

Long-term and dose-response cadmium toxicity (TAB B Toxicity Review, p. 21–30)

  • U.S. NHANES (2003–2004, CDC 2009) population biomonitoring (≥ 20 years of age): geometric mean blood cadmium 0.378 µg/L; geometric mean urine cadmium 0.260 µg/L. Females: blood cadmium 0.326 µg/L, urine cadmium 0.216 µg/L. Males: blood cadmium 0.283 µg/L, urine cadmium 0.206 µg/L (TAB B, p. 21).
  • Estimated cadmium half-life in human tissues (Kjellström and Nordberg 1978, 1985 toxicokinetic model): 6–38 years in kidney; 4–19 years in liver (TAB B, p. 23).
  • Estimated daily excretion fractions of body cadmium burden (Kjellström and Nordberg model): 0.007 percent/day in feces; 0.009 percent/day in urine (TAB B, p. 23). (TAB B contains an internal contradiction on the urine-vs-feces attribution: p. 22 of the Toxicokinetics section reports the fractions as urine = 0.007 % and feces = 0.009 %, while p. 23 of the Elimination section reports the same fractions as feces = 0.007 % and urine = 0.009 %. The wiki value above adopts the p. 23 Elimination-section attribution; see Verification notes for the source-citation evidence on both pages.) Oral absorption of ingested cadmium: approximately 1–10 percent (ATSDR 2008; TAB B, p. 22). Dermal absorption: less than 1 percent of administered dose (ATSDR 2008; TAB B, p. 22).
  • Age-related kidney/liver cadmium accumulation (Kjellström 1979): geometric mean tissue concentrations increase from less than 1 µg/g in early childhood to a peak of 40–50 µg/g in the kidney and 3–4 µg/g in the liver at around 50–60 years of age, then decrease over time after age 60 (TAB B, p. 22).
  • Järup et al. (2000) — Swedish kidney epidemiology near a Ni-Cd battery plant: 1,021 individuals (n=799 living near plant; n=222 battery workers). Mean urinary cadmium 0.81 µg/g creatinine (males) and 0.65 µg/g creatinine (females). The prevalence of abnormal pHC (human complex-forming glycoprotein, also known as α1-microglobulin) values was estimated to increase by 10 percent at a urinary cadmium level of 1 µg/g creatinine. The European Chemicals Bureau (2007) recalculation against the raw data, accounting for age differences between the reference and study populations, gave the urinary cadmium level associated with a 10 percent increased probability of abnormal pHC values (20 percent total probability) as 2.62 µg/g creatinine for the total population; in the nonworker subgroup a urinary cadmium level of 0.5 µg/g creatinine was associated with a 13 percent probability (doubling of reference-population probability) of abnormal pHC values (TAB B, p. 24).
  • Suwazono et al. (2006) — Swedish women, key study for chronic ADI: 820 Swedish women with no particular environmental or occupational cadmium exposure. Using a benchmark-dose approach and a cutoff based on the 95th percentile of urinary protein excretion estimated for a person with no cadmium exposure, the critical urinary cadmium concentration was estimated at 0.6–1 µg cadmium/g creatinine for a 5 percent excess risk based on excretion of NAG (N-acetyl-β-D-glucosaminidase) and pHC. The lower confidence limit of the BMD (the BMDL) was 0.5 µg cadmium/g creatinine (TAB B, p. 25; Staff Report, p. 4).
  • Cadmium bone-effect critical concentrations (Suwazono et al. 2010 reanalysis of Swedish women): critical urinary cadmium concentration of 1.8–3.7 µg/g creatinine for a 5 percent excess risk of low bone mineral density; BMDL of 1.0–2.1 µg/g creatinine (TAB B, p. 27).
  • U.S. women NHANES analysis (Gallagher et al. 2008, NHANES III 1988–1994 and 1999–2004): women ≥ 50 years of age with urinary cadmium between 0.50 and 1.00 µg/g creatinine had a 43 percent greater risk for osteoporosis relative to those with levels less than or equal to 0.50 µg/g creatinine. Authors estimated about 21 percent of osteoporosis prevalence among women at least 50 years of age may be attributable to cadmium (TAB B, p. 27).
  • Carcinogenicity: cadmium and cadmium compounds are listed as “known to be human carcinogens” in the U.S. National Toxicology Program Report on Carcinogens (NTP 2005), based on inhalation-exposure studies in workers. Cadmium is not classified as an oral carcinogen from oral-exposure studies (insufficient evidence in humans and experimental animals) (TAB B, p. 29).

Behavioural and ingestion-pathway data underlying the FHSA assessment (Staff Report, p. 8)

  • Emergency-room-treated injuries from ingestion of consumer products by children, 2000–2005 (CPSC analysis of NEISS): about 302,587 estimated injuries, nearly 80 percent in children under age 7 years. The most commonly swallowed objects were coins (≈ half of ingestions), jewelry, toys not elsewhere classified, and nails, screws, tacks, or bolts (CPSC 2006b, cited in Staff Report).
  • GI transit time for ingested foreign bodies in children (Macgregor and Ferguson 1998): 100 children ages 9 months to 13 years swallowed various foreign bodies including coins, ball bearings, pins, marbles, screws, buttons, a light bulb, a novelty nail file, and a clothespin. Total transit time 1 to 46 days; peak time of passage 2 days; median time 6 days. Mean transit time increased with age — greater than 15 days for 13-year-olds; typically 5 days for 4-to-10-year-olds (Staff Report, p. 8–9). This is the data underlying the 24-hour acid-extraction test duration choice.

Methods (brief)

The 109-page PDF contains (i) a 2-page transmittal letter from Colin Church (CPSC Voluntary Standards Coordinator) to Brent Cleaveland (Chairman, ASTM F-15.24 Subcommittee on Children’s Jewelry, c/o Fashion Jewelry and Accessories Trade Association), dated October 19, 2010; (ii) the Staff Report memorandum on Children’s Cadmium-Containing Metal Jewelry, dated October 14, 2010, from Kris M. Hatlelid (Toxicologist, CPSC Directorate for Health Sciences) to Mary Ann Danello (Associate Executive Director, Directorate for Health Sciences) through Lori E. Saltzman (Director, Division of Health Sciences) — 13 pages including references; (iii) TAB A: Staff Responses to Peer Review Comments, dated October 14, 2010, from Kris M. Hatlelid (5 pages), documenting CPSC staff responses to comments received from three external scientific peer reviewers in June–July 2010 on the draft Staff Report; (iv) TAB B: Toxicity Review of Cadmium, dated October 14, 2010, from Dominique J. Williams and Kris M. Hatlelid (90 pages including references and supporting tables), providing the underlying CPSC Health Sciences staff assessment of cadmium toxicity used to support the Staff Report’s exposure derivations.

No new analytical chemistry or product-content measurements are reported. The Staff Report’s derivation methodology consists of: (a) selection of Suwazono et al. (2006) as the key chronic-toxicity study from a body of European and Japanese kidney-epidemiology dose-response analyses (other dose-response studies considered in TAB B include Järup et al. 2000; Jin et al. 2004; Kobayashi et al. 2006; Shimizu et al. 2006; Uno et al. 2005); (b) PK-conversion of the Suwazono BMDL of 0.5 µg cadmium/g creatinine to a cadmium intake level of 0.33 µg/kg/day using the ATSDR 2008 European-population analysis; (c) application of a 3-fold uncertainty factor (reduced from default 10) to derive the chronic ADI of 0.1 µg/kg/day; (d) selection of Borzelleca et al. (1989) as the key acute-toxicity animal study and application of a 100-fold composite uncertainty factor (10× animal-to-human × 10× sensitive subpopulations) to the NOAEL of 1.1 mg/kg/day to derive the acute ADI of 11 µg/kg/day; (e) translation of the ADIs to per-day exposure values for a 2-to-6-year-old (18.2 kg body weight); (f) translation of the per-day exposure values to laboratory extraction-test thresholds (6-hour saline extraction → 18 µg total cadmium for mouthable items based on CPSC 2001 mouthing-observation data; 24-hour acid extraction → 200 µg total cadmium for swallowable items based on Macgregor and Ferguson 1998 GI transit-time data). The acute ADI of 11 µg/kg/day is the staff’s response to peer-review comments that the draft document’s swallowing-scenario approach was inadequate; the revised analysis uses published acute studies in experimental animals rather than the original 1973 Nordberg case-study extrapolation, which staff judged inadequately documented for derivation of a quantitative exposure limit.

The peer-review comments addressed in TAB A include (verbatim or in summary): the form of cadmium and dose route in toxicology studies and the form expected from children’s products; the renal cortex concentration threshold concept and database; whether bone effects should be considered in derivation of the ADI; the relationship between long-term and intermediate-term LOAELs and NOAELs; reconsideration of the default uncertainty factor; modelling of short-term exposure based on effects on kidney cadmium concentration; the appropriateness of the swallowing-scenario approach; the need for and form of an acute exposure limit; how acute and chronic mechanisms of action might differ; recent data on increased urinary cadmium excretion in children ages 6–11 years and in females; and whether CPSC was deriving a higher chronic ADI than the ATSDR chronic MRL or USEPA RfD (staff response: the revised CPSC chronic ADI of 0.1 µg/kg/day is the same as the ATSDR’s draft chronic MRL).

Implications

  • HMTc audit implications. This document is the U.S. federal-agency staff position underlying CPSC enforcement of cadmium content in children’s metal jewelry under the Federal Hazardous Substances Act (FHSA). For any HMTc certification work that addresses children’s jewelry or jewelry-adjacent metal accessories sold for children, the laboratory test thresholds derived in this Staff Report (18 µg total cadmium in 6-hour saline extraction for mouthable items; 200 µg total cadmium in 24-hour dilute-acid extraction for swallowable items) are the U.S. federal-agency-defined exposure-evaluation reference points. Both thresholds are the CPSC staff’s per-FHSA test method that would, if exceeded, indicate the product may meet the FHSA’s “hazardous substance” criteria. The underlying chronic ADI of 0.1 µg/kg/day (cadmium) is identical to the ATSDR (2008) draft chronic-duration oral MRL and is at the more conservative end of the international convention space (compared with the U.S. EPA RfD of 1.0 µg/kg/day for food and 0.5 µg/kg/day for water, and the WHO/JECFA PTMI equivalent to about 0.8 µg/kg/day). HMTc thresholds set in alignment with the 0.1 µg/kg/day chronic ADI should be tagged regulatory-alignment; thresholds set below should be tagged precautionary per Part 19 conventions.
  • Cross-reference to compliance baselines. The CPSC test thresholds in this Staff Report are the test-method translation of the ADI; they are not content-restriction limits in the sense of Washington State CSPA’s 40 ppm cadmium content limit on children’s products. The two regimes operate on different axes — FHSA tests extraction/migration in mouthing or swallowing scenarios, CSPA restricts substrate content. Compliance findings in Cadmium and Other Metals in Children's Jewelry (cadmium up to 98.4 % by mass in necklace pendants sold with children’s apparel) would, on per-extraction laboratory testing under the CPSC method, likely exceed the 18 µg saline or 200 µg acid extraction thresholds, pending per-item extraction testing under the CPSC 6-hour saline and 24-hour dilute-acid protocols; the FHSA and CSPA regimes are complementary, with FHSA addressing the exposure end and CSPA addressing the content end of the same hazard. The Sekerak 2016 findings are not extraction-tested in that source and are not extraction-tested in this CPSC document; the inference is a category-level cross-reference, not a CPSC-published finding.
  • App. Not directly relevant to ingredient contamination_profile data because no food-matrix occurrence values are reported. Relevant to a future children’s-product screening surface that explains the CPSC’s per-item extraction-test framework (mouthing vs swallowing scenarios) and the underlying 0.1 µg/kg/day chronic and 11 µg/kg/day acute ADIs.
  • Courses. Worked example for the regulatory-derivation module of how an agency uses a benchmark-dose epidemiological study (Suwazono et al. 2006) plus PK conversion (ATSDR 2008) plus a reduced uncertainty factor (3 rather than 10) to derive a chronic ADI for an oral exposure to a metal contaminant, and then how that ADI is translated to per-item laboratory extraction-test thresholds via mouthing-time and GI-transit-time behavioural data. The acute ADI derivation from an animal NOAEL using a 100-fold composite uncertainty factor is a complementary worked example. The peer-review-response document (TAB A) is a useful primary-source example of agency-staff responsiveness to external scientific review of a draft regulatory-derivation document.
  • Microbiome. Not applicable. No biological matrix and no exposure-pathway measurements relevant to gut microbiome interaction. The toxicokinetics discussion in TAB B is limited to absorption (≈ 1–10 % oral), tissue distribution (kidney and liver predominant), metabolism (metallothionein binding), and excretion (≈ 0.007 % feces, 0.009 % urine per day of body burden).

Update history

No substantive edit history is available in this build. The full commit record is available in git.

Referenced by