Maysson Yaghi and colleagues bought sixteen baby talcum powders from pharmacies in Benghazi in November 2024 and measured six metals by flame atomic absorption after nitric acid and hydrogen peroxide digestion. Each sample was digested in triplicate. Table 4 prints the instrument readings of the digest in ppm. Table 5 prints the same results as mg/kg of powder. For every detected cell except one, the Table 5 value is the Table 4 value multiplied by 25, which is what a 2 g sample made up to 50 mL would give. The paper does not state the final volume. The authors also model chronic daily dermal intake, hazard quotients, and cancer risk for infants aged 6 months to 3 years.
Products are coded P1 to P16. This page does not name brands.
Key numbers
Table 5, mg/kg of powder, mean of triplicate digests. LDL is printed where the metal was below detection. The limit of detection itself is not printed. Chromium is total chromium.
| Metal | Samples below detection | Minimum detected | Maximum | Mean as printed (LDL counted as zero, n = 16) |
|---|---|---|---|---|
| Al | 0 of 16 | 12.325 (P2) | 275.475 (P11) | 113.183 |
| Cd | 5 of 16 | 2.275 (P1) | 22.600 (P14) | 5.814 |
| Cr | 4 of 16 | 1.550 (P8) | 49.975 (P9) | 18.014 |
| Cu | 0 of 16 | 7.525 (P10) | 77.475 (P11) | 27.495 |
| Pb | 6 of 16 | 0.65 (P9) | 15.575 (P16) | 4.591 |
| Ni | 6 of 16 | 0.675 (P13) | 54.700 (P11) | 10.436 |
The per-metal means reproduce as the column sum divided by 16. Every per-sample cell is in the number ledger and in the structured evidence file for this source.
Two nickel results exceed 50 mg/kg: P3 at 50.125 and P11 at 54.700. Lead above 10 mg/kg appears in P3 (12.775), P12 (10.45), and P16 (15.575). Cadmium above 10 mg/kg appears in P10 (19.375), P14 (22.600), and P16 (11.425).
Values that are kept but not used
Sample P5 nickel reads 0.546 ± 0.004 ppm in Table 4 and LDL in Table 5. The Table 5 row total for P5 equals the sum of the other five metals, so Table 5 treats it as not detected. Neither figure is chosen.
The Total Metals column in Table 5 does not equal the sum of the six printed cells in eight of sixteen rows (P2, P4, P6, P8, P10, P11, P12, P16), and the printed mean total does not equal the mean of the printed totals. The per-metal cells are unaffected. The totals are not used.
The Table 4 Mean row follows no single averaging convention for cadmium, chromium, and nickel. The Table 5 means are the ones that reconcile.
The prose disagrees with Table 5 in several places. It gives the highest cadmium as 22.900 mg/kg where Table 5 prints 22.600. It describes lead as ranging from 4.500 to 15.575 mg/kg and names cadmium in that sentence, while Table 5 prints detected lead from 0.65. It says nickel was undetectable in seven samples, where Table 5 shows six. The abstract says all six metals exceeded the authors’ reference values in all powders, which Table 5 does not support for any metal except aluminium. Table values are used throughout.
The row labelled FDA in Table 5 (Al 0.400, Cd 3.000, Cr 5.000, Cu 13.000, Pb 0.100, Ni 0.600 mg/kg) is attributed in the text to references that do not set baby-powder limits: a 1989 US EPA risk guidance, a 2006 study of arsenic in gold-mine workers, and an FDA web page on cosmetic testing. This page does not use that row for any comparison.
Methods (brief)
Two grams of each powder were digested with 10 mL of 97% nitric acid at 70 to 80 °C for 30 minutes, then with 5 mL of 35% hydrogen peroxide heated to white fumes, made up to an unstated final volume with deionised water, settled overnight, and filtered through Whatman No. 42 paper. Metals were read on a GBC 932 Plus flame atomic absorption spectrometer against standards in 0.1 M nitric acid. The paper reports calibration R² above 0.99 and a repeatability criterion of relative standard deviation below 20%. It does not report limits of detection or quantification, certified reference materials, spike recoveries, or blank values.
Sampling was described as systematic random, within one production month, from different pharmacies. One product per code was analysed.
The risk model uses skin surface areas and body weights by age and sex (Table 2), reference doses (Table 3), and slope factors for all six metals. Several Table 3 and Table 6 entries break the age trend of their neighbours. The risk outputs are the authors’ model results and are not recorded as measurements.
Evidence fitness
The study supports per-product total-metal concentrations in baby talcum powders on the Libyan market, on an as-sold basis, for six metals. It does not measure arsenic, mercury, hexavalent chromium, or tin. The limitations section says four metals were evaluated and names five; six were measured.
Limitations
No detection limits are printed, so an LDL cell cannot be bounded. There is one product per code and no between-lot replication. The table-arithmetic and prose inconsistencies above are recorded in the flags file for this ingest. Several comparison values the authors cite from other studies point to reference-list entries that are unrelated papers. The corresponding author is listed as Maysson M. Yaghi (maysoon.yaghi@uob.edu.ly).
Related evidence
Update history
No substantive edit history is available in this build. The full commit record is available in git.