Findings and scope
The study measured lead in 136 brands of calcium-containing products bought in California in 1996, grouped by calcium source: natural (bonemeal, dolomite, fossil oyster shell) or synthesized and refined (nonchelated salts, mainly calcium carbonate, and chelated forms, mainly calcium citrate), plus antacids and infant formulas (PDF p.1 to p.2). Mean lead per gram of dried product ranged from 0.12 µg/g in refined antacids to 0.97 µg/g in dolomite products (Table 4, PDF p.4). The highest single brand was a refined nonchelated product at 10.05 µg/g. Lead was below the 0.02 µg/g detection limit in all six infant formulas. The authors report that two thirds of the products would fail the April 1999 California criterion of 1.5 µg lead per daily dose of calcium, and that the median lead contribution from a daily dose was 2.38 µg (PDF p.1 and p.5). Category differences were not statistically significant (analysis of variance, PDF p.3).
Key numbers
Lead in brands purchased in 1996, µg/g dry weight, as printed (Table 4, PDF p.4).
| Category | Brands (n) | Mean ± SD | Range |
|---|---|---|---|
| Bonemeal | 9 | 0.60 ± 0.39 | 0.21 to 1.38 |
| Dolomite | 5 | 0.97 ± 0.49 | 0.39 to 1.56 |
| Oyster shell | 26 | 0.88 ± 0.51 | 0.12 to 2.10 |
| Refined (nonchelated) | 37 | 0.73 ± 1.60 | 0 to 10.05 |
| Refined antacids | 20 | 0.12 ± 0.16 | 0.01 to 0.71 |
| Chelated | 33 | 0.57 ± 0.54 | 0.04 to 2.8 |
| Infant formula, powder | 3 | < 0.02 | not applicable |
| Infant formula, canned liquid | 3 | < 0.02 | not applicable |
In the interlaboratory calibration, seven unnamed products ranged from 0.06 µg/g (a refined lozenge) to 2.60 µg/g by graphite furnace AAS and 1.72 µg/g by ICP-MS (a dolomite tablet) (Table 2, PDF p.3). One lot of refined calcium carbonate powder measured 0.167 to 0.187 µg/g across four laboratories (Table 1, PDF p.3).
Methods and Evidence Fitness
Five to ten units per lot were weighed and ground; aliquots were dried at 75 °C, digested in nitric acid in trace-metal-clean Teflon labware and analysed by high-resolution magnetic sector ICP-MS with a bismuth internal standard (PDF p.2). The method detection limit was 0.02 µg/g and the limit of quantitation 0.06 µg/g, with a coefficient of variation of 1.7% and 90 to 100% recovery of NIST SRM 1486 bonemeal (Table 3, PDF p.3). Ten percent of samples were digested in duplicate and three procedural blanks ran with each batch. The authors report that graphite furnace AAS after the same digestion recovered only 51.0 ± 5.6% of reference-material lead, which is their basis for the abstract’s warning that weaker methods can underestimate lead by as much as 50%.
Evidence Fitness is Strong occurrence evidence for lead in calcium supplement categories in 1996. Quality control is well documented and the category summaries can be used as measured occurrence data. The data are historical and unbranded, and the authors expect lower lead in later products, so no row is headline-eligible for the current market.
Assessment limits
Table 4 is per gram of dried product, whereas the regulatory comparisons in the text and Figures 1 and 2 are per gram or per 800 mg of calcium. The per-calcium distributions are shown only as charts and were not captured. The abstract states that nonchelated antacids and infant formulas ranged from nondetectable to 2.9 µg lead per gram of calcium; this calcium-normalised figure is not directly comparable with Table 4. The Refined range is printed with a lower bound of 0, below the stated limit of quantitation. The basis of the daily dose behind the 2.38 µg median (800 mg or 1 g of calcium) is not stated. In the calibration, the dolomite tablet differed between laboratories by 41% (Table 2). Table 4 also reprints earlier investigators’ values; these are secondhand and were not captured.
Provenance notes
The PDF (5 pages, scanned with a text layer) was delivered in batch 05 of order HMI-GAP-INGEST-1 on 22 October 2026. The SHA-256 matches the batch manifest. The PDF does not print a DOI; it prints the citation Environ Health Perspect 108:309-313 (2000), which corresponds to manifest DOI 10.1289/ehp.00108309. The printed title matches the manifest title. The manifest metal list includes arsenic, but the paper measured only lead.
Related evidence
- Ingredient class — vitamin/mineral (single, multi-mineral, multi-vitamin/mineral)
- Infant Formula
- Contribution of Lead from Calcium Supplements to Blood Lead
- Lead
Update history
22 October 2026: Ingested from the batch-05 PDF. 34 lead records were added to the evidence register.
Update history
No substantive edit history is available in this build. The full commit record is available in git.