Heavy metals in baby lotions, oils, and powders

Abstract

Three product forms share a shelf and a body site and do not share an evidence base. Baby talcum powder is the only form in this group measured across a full metals panel, and it is now measured by two independent multi-product studies rather than one. Yaghi and colleagues measured sixteen talcum powders bought in Benghazi pharmacies by flame atomic absorption and reported, in milligrams per kilogram as sold, aluminium in every product from 12.325 to 275.475, copper in every product from 7.525 to 77.475, and, among the products where the metal was detected, chromium 1.550 to 49.975, cadmium 2.275 to 22.600, nickel 0.675 to 54.700, and lead 0.65 to 15.575. Almugren and colleagues measured four coded talcum baby powders by inductively coupled plasma mass spectrometry and reported, in micrograms per kilogram, total chromium 4883.0 to 5846.0, nickel 2082.0 to 3102.0, lead 1944.0 to 2585.0, total arsenic 63.0 to 672.6, copper 175.2 to 313.2, cobalt 34.7 to 41.7, and cadmium 3.4 to 4.7, detected in all four products for every one of those elements. That set supplies the first commercial baby-powder total-arsenic figures this synthesis can promote. A 2019 Chinese regulatory re-test of one children’s talc powder batch reported lead at 19.3 milligrams per kilogram against a 10 milligram-per-kilogram limit. Baby lotion rests on one product: sample L19 in Arshad and colleagues, a baby lotion from northern Pakistan, at lead 7.94, nickel 6.16, iron 2.01, chromium 0.67, and cadmium 0.27 milligrams per kilogram. Baby oil has no measured value in the documents opened for this synthesis. Arsenic is total arsenic and chromium is total chromium in every study here; no source speciates arsenic, mercury, or chromium, and no source measures inorganic arsenic, methylmercury, hexavalent chromium, or tin in any of these products. No source measures how much of any metal in a powder, a lotion, or an oil crosses infant skin. Separately from every commercial measurement, three case reports describe folk and traditional powders used on or near infants, at lead from 180.1 milligrams per kilogram in one traditional topical powder to 62 to 64 percent by weight in a folk diaper powder, the latter being four to five orders of magnitude above any commercial product here; those are a distinct hazard class and are never pooled with commercial occurrence. This synthesis states what those documents support. It does not rank products now on the shelf, and it does not set a certification limit.

Product anatomy

The three forms are built differently, and the difference decides which metals are plausible before any is measured.

A baby powder is a milled mineral or a milled starch, sold dry. The talc form is finely ground hydrated magnesium silicate, valued for slip and moisture absorption. Aluminium, silicon, magnesium, iron, and trace metals travel with the mineral from the deposit, so a talc powder’s metal load is largely the geology of its mine and the depth of its refining, not an additive. The starch form is milled corn starch or another plant starch, sometimes with kaolin, and carries a plant rather than a mineral background. That distinction is not cosmetic. The Shenzhen Consumer Council test discussed below found that the powders exceeding the German orientation values were the plant-starch powders without talc, which is the opposite of the naive expectation that the mineral is always the dirtier base. The mineralogical hazards of talc, asbestiform contamination and pulmonary effects, are a separate question from heavy-metal content and are not this paper’s subject.

A baby lotion or cream is an emulsion: water, an oil or petrolatum phase, an emulsifier, a preservative system, and often fragrance. A baby oil is a single-phase oil, usually light mineral (paraffin) oil, sometimes a plant oil, with fragrance. In an emulsion or an oil the metal does not arrive with a milled mineral. It arrives with a raw material, a colourant, a pigment in a tinted cream, or a contaminated process water, and it is distributed through a matrix that is mostly water or mostly oil. A total digest of the finished product cannot say which input carried the metal.

The exposure surface is the same for all three: infant skin, often on the diaper area, sometimes broken or inflamed. A powder is also inhalable during application, which a lotion and an oil are not. None of these routes is quantified for a metal in the documents opened here.

Measured evidence

The measured evidence divides cleanly by form. Powder is where the numbers are; lotion is one product; oil is a gap.

The values printed below are the finished-product concentrations that this ingest promoted as uncontested, in the native basis each source reported, of product as sold. Nothing is pooled across sources, and the two talcum-powder sets are reported side by side rather than combined. Yaghi and colleagues reported in milligrams per kilogram, equivalently parts per million. Almugren and colleagues reported in micrograms per kilogram, equivalently parts per billion, and that unit is carried through as printed; dividing an Almugren figure by one thousand gives the milligram-per-kilogram equivalent exactly, and no other arithmetic is applied to it. Values that a source’s own tables contradict, that carry an unresolved flag, or that are reported on a basis that cannot be reconciled with as-sold product are held out of this table and named in the inventory that follows, so the reader can see why they are not summarised. Detection counts are the source’s own; a below-detection cell is not a zero, because most of these studies printed no detection limit.

Two further points govern how the table should be read. The two talcum sets do not agree in magnitude, and the disagreement is largest where it matters most: Almugren’s four powders sit near 2 to 2.6 milligram-per-kilogram lead with cadmium roughly three orders of magnitude below Yaghi’s detected cadmium, while Almugren’s total chromium at roughly 4.9 to 5.8 milligrams per kilogram sits inside Yaghi’s detected chromium range. Different markets, different instruments, and four products against sixteen are all sufficient to explain that spread, and no source here reconciles it, so neither set is presented as the truer one. Separately, every figure in this table is a commercial product. The folk and traditional lead-oxide powders described later in this file are excluded from the table entirely and have no cell in it.

Baby oil is the form with no measured value in this evidence. The single product ever measured as a baby oil in the documents opened here is one paraffin-oil sample in Almukainzi and colleagues, and every one of its cells is contested for the reason given in the inventory below, so no baby-oil value is carried into the table. No cooking-oil or adult-product number is substituted for the gap. A mustard-oil dataset and an adult-cosmetic oil are present in the corpus; neither is a baby-oil measurement, and the mustard-oil paper makes no skin-application claim, so they do not fill this row.

Source-by-source inventory

Yaghi and colleagues measured aluminium, cadmium, total chromium, copper, lead, and nickel by flame atomic absorption in sixteen baby talcum powders bought from Benghazi pharmacies in November 2024, each digested in triplicate. Their per-product Table 5 cells, in milligrams per kilogram as sold, are the promoted values summarised above. The study printed no detection limits, reference materials, or recoveries; it reported one product per code; and its own tables carry internal disagreements that keep several of its other numbers out of this synthesis. The Total Metals column does not add up in half the rows, the Table 4 digest-reading means are a separate basis, the prose maxima differ from Table 5, sample P5 nickel is detected in one table and below detection in the other, and the row the authors label as FDA limits cites references that do not set such limits. None of those disputed figures is printed here. Arsenic, mercury, and tin were not in the study’s panel ¹.

Almugren and colleagues measured four coded commercial talcum baby powders by inductively coupled plasma mass spectrometry and printed the results in micrograms per kilogram with standard deviations. Every element in the panel was detected in all four products: total chromium 4883.0 to 5846.0, nickel 2082.0 to 3102.0, lead 1944.0 to 2585.0, total arsenic 63.0 to 672.6, copper 175.2 to 313.2, cobalt 34.7 to 41.7, and cadmium 3.4 to 4.7. Those cells are promoted. This is the second multi-product commercial talc set in the corpus and the only one that includes arsenic, so it is what moves the arsenic row off no data found. Three limits sit on the paper that do not travel with it. The authors state that a Malaysian authority and the ASEAN Cosmetic Scientific Body permit 5 parts per million arsenic as an impurity, and they attribute a 3 part-per-million arsenic figure to the United States Food and Drug Administration; those are the authors’ citations, they were not checked against the primary instruments during ingest, and they are not entered as limits anywhere in this file. Two mechanical caveats are recorded rather than smoothed: the standard deviation on one chromium cell is line-wrapped in the text layer and was reconstructed, and the arsenic average the paper prints, 307.8 micrograms per kilogram, is not the arithmetic mean of the four arsenic cells, which is 228.25, so the printed average is left as the authors’ sentence and no average is used as a headline here. The paper also reports naturally occurring radioactive material, which is outside this synthesis. Aluminium, mercury, and tin were not in the panel, and n equals four ².

A 2019 Chinese national non-compliance notice, relayed by a regulator-affiliated news site, named a children’s talc powder batch at 19.3 milligrams per kilogram lead on re-test against a 10 milligram-per-kilogram limit, and a provincial decision found three batches of that powder over the limit, all made from the same talc raw material. That single re-test lead value is promoted as a regulatory measurement. It is one batch, not a survey, and the method behind the re-test is not printed in the news relay ³.

Arshad and colleagues analysed sixty-three cosmetics from northern Pakistan in triplicate by flame atomic absorption, with a NIST standard reference material recovery of 97 to 102 percent. One product, sample L19, is a baby lotion, and its five detected cells, lead 7.94, nickel 6.16, iron 2.01, chromium 0.67, and cadmium 0.27 milligrams per kilogram, are the promoted baby-lotion values. It was the second-highest lead and cadmium among the thirty lotions in that study. A petroleum jelly in the same set contained lead 1.49 and nickel 2.35 milligrams per kilogram, with cadmium and chromium below detection; it is a jelly, not a baby lotion, and is context rather than a promoted baby-lotion value. Two of the study’s lotion summary rows disagree with its per-sample tables for chromium and nickel and are not used ⁴.

The Shenzhen Consumer Council tested twenty infant powders, both talc-based and plant-starch-based, and reported through the press that all met the Chinese national standard. The maxima across all twenty samples were lead 3.5, arsenic 0.87, and cadmium 0.31 milligrams per kilogram, and the four samples that exceeded the German orientation values were plant-starch powders without talc, one of which exceeded all three. These three maxima are promoted only as grey literature: a news-reported compliance test whose per-sample data and analytical method were not obtained. They are treated here as the only figures the corpus carries for a plant-starch baby powder, and they are deliberately kept out of the abstract and out of every summary comparison as a headline number ⁵.

Omenka and Adeyi measured zinc, cadmium, lead, and nickel by flame atomic absorption in thirty-five personal care products from Ibadan markets, including two baby powders with ash above 95 percent and four petroleum jellies labelled for babies. One baby powder contained 468 milligrams per kilogram lead, a magnitude two other laboratories reproduced at 508.5 and 364.63; the other was reported at 36.3 milligrams per kilogram cadmium, which a third laboratory reported as 0.38. The paper reports on dried sample mass but never states its reporting basis, and for creams and jellies the drying correction can be large, so none of the Omenka values is promoted or summarised above. They are recorded because the 468 milligram-per-kilogram lead result, corroborated across laboratories, is the largest commercial baby-powder lead value in the corpus and would understate the hazard if it were dropped silently rather than held for a stated basis reason ⁶.

Almukainzi and colleagues measured eleven elements by inductively coupled plasma mass spectrometry in twenty-one topical cosmetics from Riyadh, including one baby oil of light paraffin oil and perfume and one baby powder of talc, zinc oxide, and salicylic acid. The reported baby-oil means were aluminium 26.81, total chromium 21.17, manganese 2.62, iron 193.16, cobalt 0.08, nickel 13.16, copper 41.72, zinc 71.68, total arsenic 0.15, cadmium 0.23, and lead 5.52 parts per million; the baby-powder means included lead 23.18 and aluminium 3774.14 parts per million. In nearly every cell the standard deviation is about 1.41 times the mean and the replicate count is not stated, so these means are contested and none is promoted. They appear here, and only here in the body, as a labelled alternative: they are the sole baby-oil measurement in the corpus, and they are not carried into the abstract, the measured-evidence table, or any comparison, because their dispersion is unresolved ⁷.

A Washington State Department of Ecology survey analysed one baby powder of unstated formulation, printing total chromium 2.2, zinc 2.3, cobalt 0.4, copper 0.4, lead 0.4, and molybdenum 0.2 parts per million. The survey’s quality plans name a 1.0 part-per-million reporting limit for antimony, arsenic, cadmium, lead, and, where listed, total chromium. Cobalt 0.4, copper 0.4, lead 0.4, and molybdenum 0.2 sit below 1.0 part per million; total chromium 2.2 and zinc 2.3 do not. The report does not define its 0.0 entries. Two further baby-labelled samples in that survey are the containers of a baby cream and a petroleum jelly, listed as packaging, so their values describe the packaging and not the contents. Because the 0.0 convention and the reporting limits are unresolved for this survey, the baby powder values are not promoted ⁸.

A 2025 community-funded test reported one lot of one United States baby lotion as non-detect for lead, cadmium, and mercury below 5 parts per billion and arsenic below 10 parts per billion. The laboratory report is not in the published copy. It is grey literature, it is a set of non-detect thresholds rather than measured concentrations, and it is not promoted or summarised ⁹.

Rosnan reported eight talcum powders from Malaysia in milligrams per litre of digest solution, a basis that cannot be converted to a product concentration from the excerpt available and that is not stated to be baby powders. Those digest readings are not promoted and are not summarised ¹⁰.

Rehman and colleagues digested thirty talcum powders bought in Faisalabad and measured lead, cadmium, cobalt, total chromium, and copper by atomic absorption spectrophotometry, reporting coded means from P1 to P30 in parts per million. Across the thirty means, lead runs from 0.001 to 1.053, cadmium from 0.001 to 0.08, cobalt from 0.001 to 0.18, total chromium from 0.083 to 0.35, and copper from 0.079 to 0.35. None of those values is promoted, for two independent reasons. The digest was diluted to a volume the paper calls the desired volume without stating it, so the printed parts per million cannot be confirmed as milligrams per kilogram of powder; and the table is not split between baby powder and adult dusting powder, so no subset of it can be attributed to a baby product. The paper’s own abstract ranges disagree with its Table 1 for lead and cobalt, and its abstract claim that lead was extremely high in all brands is not what the table shows, since the chromium and copper means often exceed the lead means. Safe-limit figures the paper attributes to the Food and Drug Administration and to unnamed sources are recorded on the source page as author statements and are not entered as limits. This set is the largest talcum sample count in the corpus and is noted for that reason, not summarised ¹¹.

A conference abstract reports bulk X-ray fluorescence on one hospital-designated talcum baby powder bottled in 1985, five replicate samples each analysed ten times, at approximately 2000 parts per million nickel and approximately 500 parts per million total chromium, with iron described as a few weight percent. Electron microscopy placed chromium in chromite or chromium-bearing magnetite crystals a few micrometres across and placed nickel and iron inside platy talc particles and talc fibres, which the authors read as consistent with an ophiolitic or ultramafic talc source. Those two figures are the highest nickel and total chromium ever reported for a baby powder in this corpus, roughly thirty-six times the highest detected Yaghi nickel and ten times the highest detected Yaghi chromium. They are not promoted and are deliberately excluded from every range and every market statement in this file, for reasons that compound: it is a meeting abstract rather than a full paper, the concentrations are stated as approximate with no digestion method, calibration, or detection limit printed, it is one bottle, and that bottle was filled in 1985 and so describes neither a current product nor a current deposit. It is recorded because a single mineralogically characterised bottle at that magnitude is a reason to measure nickel in talc against its geological source, and because dropping it silently would understate what the upper tail of a talc-sourced powder can hold. Chromium is not speciated ¹².

Nnorom measured lead, cadmium, cobalt, and total chromium by flame atomic absorption in fifty-six talcum powders from southeast Nigeria, reporting overall means of lead 5.0 plus or minus 1.0 milligrams per kilogram with a range of 0.4 to 41, cadmium 2.1 plus or minus 0.3 with a range from non-detect to 8.1, cobalt 0.7 plus or minus 0.1 with a range from non-detect to 2.4, and total chromium 0.2 plus or minus 0.1 with a range from non-detect to 2.7. Two products exceeded 10 milligrams per kilogram lead, and excluding those two the lead mean was 3.7 plus or minus 0.3 with a range of 0.4 to 8.7. This study is adjacent rather than central: the set is predominantly adult facial and body talc, the paper summarises baby powders only as a group of three, and the source page does not transcribe that subgroup’s own concentrations, so no baby-powder figure can be drawn from it and none is stated here. The overall statistics are recorded as the nearest large-sample talc background in the corpus. Results are aggregated by manufacturer code rather than by product, no detection limit is printed so non-detects cannot be converted, chromium is not speciated, and copper was not measured despite a copper limit being cited ¹³.

Folk and traditional lead-oxide powders, a separate hazard class

Three case reports in the collection describe folk or traditional powders used on or near infants. None of them is a commercial baby product, none receives a row in the measured-evidence table above, and none is pooled with any commercial value anywhere in this file. The hazard here is a lead compound present in the preparation itself, not trace contamination of a milled mineral, and it is a poisoning mechanism rather than an occurrence distribution.

The three do not sit at one magnitude, and the difference matters. Two of them are deliberate lead-oxide preparations at 62 to 64 percent and at roughly 21 percent lead by weight, four to five orders of magnitude above any commercial powder in this file. The third is a traditional topical powder at 180.1 milligrams per kilogram, which is high against a commercial product but only about an order of magnitude above the highest promoted commercial figure here, and it was applied to an adult rather than to the infant. Grouping all three as one class is correct as to mechanism and wrong as to scale, so the concentrations are given individually below and no single multiplier is offered for the class.

A traditional Chinese topical powder used by a breastfeeding mother on her own skin contained 180.1 milligrams per kilogram lead, and her exclusively breastfed infant’s blood lead rose above 400 micrograms per litre before chelation. The powder was not applied to the infant ¹⁴.

A non-commercial Malaysian folk diaper powder used on a United States infant was 62 to 64 percent lead by weight, that is 620,000 to 639,500 milligrams per kilogram, primarily as lead oxide, with antimony at 1005, bismuth at 110, arsenic at 55, and thallium at 31 milligrams per kilogram. The arsenic is an elemental total and the paper does not speciate it. The infant, aged nine months, had venous blood lead of 13 to 18 micrograms per decilitre against a reference level below 5, and 8 micrograms per decilitre four weeks after use stopped. This full text also supplies the first bioaccessibility figures in this synthesis, from a two-hour extraction: of the lead in the powder, 39 percent was released into a simulated gastric fluid, 0.03 percent into a simulated lung fluid, and 0.01 percent into a simulated urine. The same extraction released 62 percent of the antimony, 50 percent of the arsenic, 88 percent of the bismuth, and 12 percent of the thallium into the gastric fluid. Mercury in those leachates was below 0.005 nanograms leached per kilogram of solid, which is a leachate result and not a measured mercury concentration in the powder. Two further figures in the paper, 105,000 milligrams of bioavailable lead per kilogram of powder and the statement that one milligram of powder would deliver over ten micrograms of lead, are the authors’ calculation from the gastric result rather than a second measurement, and are attributed as such. The antimony figure in the paper’s abstract is 1000 rather than the 1005 printed in the body, and the body figure is used here ¹⁵.

The third case is the one that changes the shape of this class. Two children in Guangzhou had blood lead of 330 and 248 micrograms per litre at first testing and 303 and 385 micrograms per litre at the hospital before chelation, falling to 118 and 88 micrograms per litre after treatment with calcium disodium edetate. The powder used in the home was talcum mixed with a pharmacopeial red lead preparation, lead tetroxide, and measured above 100,000 milligrams per kilogram lead by portable X-ray fluorescence and 214,000 milligrams per kilogram by flame atomic absorption. What distinguishes this report from the other two is retail availability rather than concentration: investigators visited forty-eight drugstores across fourteen towns and found forty-one of them selling a similar powder, then purchased and measured thirty-five samples. The reported averages were 817,000 milligrams per kilogram lead across sixteen samples of the red lead preparation itself and 207,000 milligrams per kilogram across thirteen powders sold as medicated baby powders. A one-off household remedy and a preparation stocked by forty-one of forty-eight drugstores are different public-health objects, and the second is the one this report documents. The same paper measured six talcum samples at an average of 5 milligrams per kilogram lead; that figure is not promoted as commercial occurrence, because the source is a case report whose own evidence-fitness statement rules out a retail talcum distribution, and because no range or per-sample result accompanies the average. A Chinese cosmetic limit below 40 milligrams per kilogram that the authors compare against was not read from the primary instrument during ingest and is not entered as a limit here ¹⁶.

Toxicology

Toxicology here is the hazard each element is known for and what, if anything, is known about it on infant skin from a powder, a lotion, or an oil. A classification or a reference value derived from ingestion or inhalation is not a dermal dose. Where a dermal number was not in a document opened for this paper, the entry says so.

Infant skin is the exposure surface, and this paper does not assign it a number. No study opened here measures the fraction of lead, nickel, chromium, or any other metal that crosses intact or inflamed infant skin from a powder, a lotion, or an oil. The barrier is more permeable in early infancy, the diaper area is occluded and often macerated, and an infant’s skin surface is large relative to body weight; these are reasons to measure the dermal route, not a measured fraction. A powder adds an inhalation route during application that a lotion and an oil do not, and that route is likewise unquantified for a metal here.

Lead is a developmental neurotoxicant, and the endpoint that dominates infant assessment is harm to the developing nervous system, with no opened threshold below which a dermal dose would be called without effect. The measured evidence is a concentration, not that dose. The commercial baby-powder lead figures run from below detection to 15.575 milligrams per kilogram in the Yaghi survey, 1944.0 to 2585.0 micrograms per kilogram across the four Almugren powders, and a single 19.3 milligram-per-kilogram regulatory re-test; the one baby lotion is 7.94 milligrams per kilogram. The corpus also holds a corroborated 468 milligram-per-kilogram lead value in a commercial baby powder, held out of the summary only for a basis reason. The two folk lead-oxide powders are four to five orders of magnitude above all of that, and a third traditional powder is about an order of magnitude above the highest promoted commercial figure; all three belong to the separate class described above. No study here measured blood lead attributable to a commercial powder, a lotion, or an oil, or the share that crosses skin from any of them. The one place this file now carries a measured release fraction for lead is the folk powder, where 39 percent of the lead entered a simulated gastric fluid against 0.03 percent into a simulated lung fluid; that is an ingestion and inhalation contrast in a lead-oxide preparation, and it is not a dermal fraction and not a commercial product.

Cadmium’s critical long-term endpoint in dietary assessment is kidney tubular dysfunction, from an oral, cumulative record that does not transfer to a dermal dose by analogy. Cadmium was detected in eleven of the sixteen Yaghi talcum powders, from 2.275 to 22.600 milligrams per kilogram, in all four Almugren powders at 3.4 to 4.7 micrograms per kilogram, and in the single baby lotion at 0.27 milligrams per kilogram. Those two talc sets differ by roughly three orders of magnitude on this element, which is the widest disagreement between them and is unresolved by any source here. No dermal-absorption fraction from any of these products was opened.

Nickel is the most common cause of metal allergic contact dermatitis, and sensitisation is acquired through skin contact and later elicited by it. The clinical literature expresses both a sensitisation and an elicitation threshold as release per unit area over time, not as a bulk concentration in a powder or a lotion. The measured evidence here is bulk content: nickel to 54.700 milligrams per kilogram in the Yaghi talcum set, 2082.0 to 3102.0 micrograms per kilogram in all four Almugren powders, and 6.16 milligrams per kilogram in the one baby lotion. A single 1985 hospital-designated bottle was reported at approximately 2000 parts per million by a meeting abstract, which is the highest baby-powder nickel figure in the corpus and is excluded from these ranges for the reasons given in the inventory. No study here measured nickel release into sweat or a simulant from any of these products, so none of the content values can be read as an elicitation risk.

Chromium toxicity depends on oxidation state. Trivalent chromium is poorly absorbed by mouth; hexavalent chromium is the sensitiser and the carcinogenic species in the occupational record. Every chromium value here is total chromium: to 49.975 milligrams per kilogram in the Yaghi talcum set, 4883.0 to 5846.0 micrograms per kilogram in all four Almugren powders, 0.67 milligrams per kilogram in the one lotion, and approximately 500 parts per million in the excluded 1985 bottle. No study speciated chromium or measured hexavalent chromium in these products. A total-chromium number cannot be read as hexavalent chromium. One mineralogical observation is worth carrying: in the 1985 bottle, electron microscopy placed the chromium in discrete chromite or chromium-bearing magnetite grains rather than dispersed through the talc, which is a reason to expect chromium in a talc powder to vary with deposit and with how thoroughly heavy minerals were separated during refining.

Arsenic hazard in regulatory work isolates the inorganic species, which carries the carcinogenic and developmental concerns. This synthesis now has one promoted commercial total-arsenic figure, from the four Almugren talcum powders at 63.0 to 672.6 micrograms per kilogram, detected in all four. The spread across those four cells is roughly tenfold. Within that set arsenic does not move with the panel as a whole: the product carrying the highest arsenic also carries the highest lead and the highest cobalt, but the lowest nickel and the lowest total chromium of the four. On four products that pattern is an observation and not a correlation, and it is recorded only as a reason not to treat one element in a talc panel as a proxy for another. No source speciates arsenic in any product in this group, so there is still no inorganic-arsenic figure at all, and the promoted range cannot be read against an inorganic-arsenic limit. The 55 milligram-per-kilogram arsenic in the folk diaper powder is a total, is three orders of magnitude higher, and belongs to the separate class above.

Mercury’s neurodevelopmental concern attaches to methylmercury after ingestion, and elemental and inorganic mercury have different kinetics. No commercial baby-powder, lotion, or oil mercury value cleared the holds here, and no study measured methylmercury.

Aluminium is the element the talcum-powder measurements find in the largest amount, detected in every one of the sixteen Yaghi products from 12.325 to 275.475 milligrams per kilogram, consistent with aluminium being a major constituent of the aluminosilicate mineral rather than a trace contaminant. Aluminium neurotoxicity and bone toxicity are documented at high systemic exposures, particularly where the kidney cannot clear the metal, and that record is oral and parenteral, not dermal. No study here measured aluminium crossing infant skin from a powder, and no promoted aluminium value exists for baby lotion or baby oil.

Copper is an essential element, and its interest here is as a marker of the mineral background rather than a primary toxicant at these concentrations; it was detected in every Yaghi talcum product from 7.525 to 77.475 milligrams per kilogram and in all four Almugren powders at 175.2 to 313.2 micrograms per kilogram. Iron, similarly, appears as a single 2.01 milligram-per-kilogram value in the one baby lotion, and is described as a few weight percent in the excluded 1985 bottle; it is not a primary hazard at the lotion level.

Cobalt is a skin sensitiser and a frequent co-reactant with nickel in patch testing, and it enters this synthesis for the first time. It was detected in all four Almugren talcum powders at 34.7 to 41.7 micrograms per kilogram, a narrow range across the four. No study here measures cobalt release from a powder or a lotion, and no instrument opened for this file sets a finished-product cobalt number for a baby product, so cobalt is a measured content with neither a release figure nor a limit to read it against.

Exposure framing

A dermal exposure from any of these products has at least four inputs: how much metal is in the product, how much of it reaches skin, how long it stays, and what fraction is absorbed. This paper assigns none of them a number, because no primary study opened here supplies, together, a product mass applied, a contact time, and a dermal-absorption fraction for any metal in a baby powder, lotion, or oil. An illustrative dose built by inventing the missing inputs would read as a risk estimate and would not be one, so none is offered. The gap is the finding.

The first measured release fractions in this file do not close that gap. The folk diaper powder was extracted for two hours into three simulated fluids, releasing 39 percent of its lead into a gastric fluid, 0.03 percent into a lung fluid, and 0.01 percent into a urine. Three things follow, and a fourth does not. The gastric figure shows that a lead-oxide powder near the mouth or on a hand is an ingestion problem more than an inhalation one, by a factor of roughly a thousand between those two fluids. The ordering also differs by element, since antimony released 62 percent and thallium 12 percent into the same gastric fluid, so a single bioaccessibility number cannot stand in for a panel. And an in vitro extraction is a laboratory fraction rather than an absorbed dose in a child. What does not follow is any transfer of these percentages to a commercial talc or starch powder or to the dermal route. The material extracted was a deliberate lead-oxide preparation, mineralogically unlike milled talc, and none of the three fluids is skin.

What can be said without arithmetic is qualitative. The measured signal in this group is concentrated in one form, talcum powder, and it is now supported by two independent multi-product sets that agree on detection and disagree on magnitude. Aluminium and copper were found in every Yaghi product, and every element in the Almugren panel was found in all four of its products, so detection in commercial talc is close to universal for the metals these panels cover while the concentrations are not settled. Baby lotion rests on a single measured product. Baby oil has no measured value at all. The powder route adds inhalation during application; the lotion and oil routes do not. Until the dermal and inhalation fractions are measured for a commercial product, a concentration in a powder is a concentration in a powder.

Regulatory landscape

These products are not one legal object. A product applied to infant skin for cleansing, moisturising, or moisture absorption can fall inside a cosmetic definition; in Canada a corn-starch or kaolin diaper-rash powder can instead be a natural health product, and a mineral-oil or petrolatum diaper-rash product a non-prescription drug. Limits are cited below only in the unit and the legal character of the instrument. The measured concentrations above are not scored against them as exceedances, because the promoted values sit on an as-sold total-content basis while several instruments are extractable limits, orientation values, or guidance rather than statutory finished-product maxima, and because the pre-statute samples predate the newer instruments.

Health Canada’s guidance on heavy-metal impurities in cosmetics treats lead above 10, arsenic above 3, cadmium above 3, mercury above 1, and antimony above 5 parts per million in a finished cosmetic as technically avoidable, subject to case-by-case enforcement. It is guidance, not a numeric statutory maximum, and it reaches baby lotions, oils, and powders sold as cosmetics.

Health Canada’s Quality of Natural Health Products Guide sets finished topical-product limits of arsenic 3, cadmium 3, lead 10, total mercury 1, and antimony 5 parts per million, which reach corn-starch and kaolin diaper-rash powders made only with natural health product ingredients. The guide’s methylmercury value is an oral limit and does not govern these topical rows.

The German BVL orientation values for cosmetics in general are lead 2.0, cadmium 0.1, mercury 0.1, arsenic 0.5, and antimony 0.5 milligrams per kilogram as sold, and baby powder was one of the monitored product groups behind them. They are orientation values under the European cosmetics regulation, not statutory maxima.

Washington State prohibits the sale of a cosmetic with intentionally added lead or lead at or above 1 part per million from 1 January 2025, and prohibits intentionally added mercury with no numeric value. The statute’s cosmetic definition does not name baby powder specifically, and the measured values here are not called violations, because the sampling basis may not match the article the statute governs and the samples are older than the statute.

Two talc specifications in the collection are for a different purpose and a different basis. The United States color-additive regulation for talc used in drugs sets lead at 20 and arsenic at 3 parts per million measured in a 0.5 normal hydrochloric-acid extract, which is extractable content and never directly comparable with a total digest, and it is a raw-material specification rather than a finished baby-powder limit. The JECFA food-additive specification for talc sets lead at 2 milligrams per kilogram for food-grade talc for oral use, which is a food-additive specification and is excluded from any topical-product comparison. Neither is a finished baby-powder limit, and neither is scored against the measured powders here.

No instrument opened here sets a finished-product numeric maximum for aluminium, copper, iron, tin, hexavalent chromium, inorganic arsenic, or methylmercury in a baby powder, lotion, or oil. Where a government has no number for a metal in these products, that gap is left as a gap rather than filled by borrowing a limit written for a different article or a different basis.

Implications

For parents

A parent can take four statements from this file and should not take a fifth. Lead, cadmium, total chromium, nickel, and copper are found in nearly every commercial baby talcum powder that has been measured across a full panel, at the concentrations listed above, and total arsenic and cobalt were found in all four of the products where they were looked for. Baby lotion has been measured in only one product in this evidence, so it cannot describe the lotion market. Baby oil has not been measured at all in the documents opened here. Tin, inorganic arsenic, methylmercury, and hexavalent chromium were not found because they were not measured or not speciated, not because a survey showed them to be absent.

The statement not to take is a risk verdict. No study here follows any of these products from the container onto infant skin and into the body, so a concentration is not a dose, and stopping or continuing a product is a caregiving decision this paper does not make.

One distinction is worth more than the rest of this section. Folk and traditional powders containing a lead compound, applied on or near infants, are a poisoning hazard of an entirely different kind from trace contamination of a commercial product. The two lead-oxide preparations in this file are four to five orders of magnitude above every commercial measurement in it, and a third traditional powder is about an order of magnitude above the highest commercial figure here. They are not the powders on an ordinary shelf. One of the three reports found such a preparation stocked by forty-one of forty-eight drugstores visited in one region, so being sold in a pharmacy or a drugstore is not by itself an assurance about what a powder contains.

For manufacturers

Talc’s metal load is largely its mine and its refining. The dominant lever on a talcum powder is deposit selection and the depth of washing, grinding, and air-floating, because aluminium, copper, and the trace metals travel with the mineral rather than being added. A starch powder carries a plant background instead, and the Shenzhen result is a reminder that a talc-free reformulation is not automatically the cleaner one on every metal.

The mineral-host observation sharpens that lever into a specific instruction. A single mineralogically characterised bottle placed its chromium in discrete chromite or chromium-bearing magnetite grains a few micrometres across, and its nickel and iron inside the talc particles and fibres themselves, at bulk concentrations far above any other baby powder in this corpus, with the authors reading the chemistry as an ophiolitic or ultramafic source. Those two hosts respond to different controls. Discrete heavy-mineral grains are in principle separable by gravity or magnetic beneficiation, whereas nickel held in the talc lattice is not removable by any amount of physical cleaning and can only be addressed by choosing a different deposit. A specification written only as a finished-powder maximum will not distinguish those cases, so a supplier qualification that asks for the deposit’s host geology alongside a certificate of analysis is doing work that the finished number cannot do. That inference rests on one historical bottle reported in a meeting abstract and should be treated as a hypothesis to test on current material, not as a characterisation of any deposit now in use.

A lotion and an oil need the metal traced to an input. A single total digest of an emulsion or an oil cannot say whether a metal came from a raw oil, a pigment, a preservative, or process water, and the finished-product number alone will not guide a corrective action. Report the finished product in milligrams per kilogram as sold, keep the digest solution as a method detail rather than the only reported unit, and record the recovery and the detection limit that most of the studies here omitted.

A bulk content number is not a sensitisation test. A laboratory that wants to speak to allergic contact dermatitis from nickel or chromium in a lotion or a powder has to measure release into sweat or a simulant over a stated area and time, not the bulk metal in a digest.

For regulators

The legal character of each number should stay visible. A Canadian guidance figure, a natural-health-product topical limit, a German orientation value, a Washington lead statute, a drug-talc acid-extract specification, and a food-additive talc specification are not interchangeable, and none of them was written as a score for a total digest of a finished baby product. Where a government has no number for a metal in these products, the gap should stay a gap; borrowing a food or a drug-raw-material talc limit to fill a finished-cosmetic cell would present a policy choice as a measurement.

Aluminium is the awkward cell for powders. It is the element the talcum measurements find in the largest amount, it is intrinsic to the mineral rather than an adventitious contaminant, and no instrument opened here sets a finished-product number for it. A regulator reading only the four food-metal panels will not see it. Cobalt is now a second cell of the same shape: measured in every product of the one panel that looked for it, a recognised skin sensitiser, and without a finished-product number in any instrument opened here. Speciation rules that already exist for food should not be relaxed because the matrix is a powder or a lotion: total arsenic is not inorganic arsenic, total mercury is not methylmercury, and total chromium is not hexavalent chromium.

Two of the new sources carry limits attributed to regulators that this file does not adopt, and the pattern is worth naming because it recurs. One paper attributes a 3 part-per-million arsenic figure to the United States Food and Drug Administration and a 5 part-per-million figure to a national authority and a regional cosmetic body; another calls 20 parts per million a lead safe limit by the same agency and gives unattributed safe-limit ranges for four more elements. Those are the authors’ sentences, none was verified against a primary instrument during ingest, and none is entered as a limit anywhere in this synthesis. A limit that enters the literature by citation and is then re-cited from paper to paper acquires the appearance of an instrument without ever being one, and a regulator comparing a measured powder against such a figure would be scoring it against nothing.

The drugstore finding belongs to enforcement rather than to limit-setting. A preparation containing a pharmacopeial lead compound, sold as a medicated baby powder in forty-one of forty-eight drugstores visited in one region, is not a contamination problem that a finished-product maximum addresses. It is a question about which preparations may be sold for application to infants at all, and it sits outside the cosmetic impurity framework that the rest of this section describes.

For researchers

The most useful next studies are specific. For baby oil, any properly reported measurement would be the first: a paraffin-oil and a plant-oil product, digested and reported in milligrams per kilogram as sold, with a stated replicate count and recovery, so the single contested sample in the corpus is no longer the only one. For baby lotion, more than one product, so the category no longer rests on a single sample. For talcum powder, surveys that print detection limits, reference materials, and recoveries, and that separate talc from starch bases. Across all three, speciate arsenic, mercury, and chromium, add tin and aluminium to the panel, and run a sweat or simulant release in parallel with total content, because the sensitisation questions are release questions. Until those studies exist, the synthesis has to keep repeating that the forms in this group are unevenly measured and that a total content on an as-sold basis is not a dermal dose.

The sources pose three questions rather than answer them. The first is why two competent talcum studies disagree by roughly three orders of magnitude on cadmium while overlapping on chromium; a study that measured powders from both markets on one instrument with one certified reference material would settle whether that is geology, method, or a reporting-basis artefact, and it should be run before either range is treated as a market estimate. The second is whether the nickel and chromium magnitudes in a single 1985 bottle appear in any current product, which requires bulk chemistry paired with electron microscopy on present-day powders of known deposit, reported as a full paper rather than an abstract, and which would also test whether nickel sits in the talc lattice or in separable heavy minerals. The third is whether the mineral form drives bioaccessibility: the only release fractions in this file come from a lead-oxide folk preparation, and running the same three-fluid extraction on a milled commercial talc would show whether trace lead in an aluminosilicate behaves anything like lead oxide, which nothing here establishes. A dermal extraction on infant-relevant skin remains the gap that none of these three would close.

Data gaps and research agenda

No dermal-absorption fraction for any metal from a baby powder, lotion, or oil was opened. Baby oil has no promoted measured value; the only baby-oil measurement in the corpus is a single contested sample. Baby lotion rests on one measured product. Tin in any of these products: no data found. Total mercury in any commercial product in this group: no promoted value. Inorganic arsenic, methylmercury, and hexavalent chromium in any of these products: no data found, and no source speciated arsenic, mercury, or chromium. Aluminium is measured in only one of the two talcum sets, because it was not in the other’s panel. Commercial total arsenic and cobalt rest on one four-product set each. Corn-starch and other plant-starch baby powders are represented only by three grey-literature maxima from a news-reported compliance test with no per-sample data and no printed method. A large corroborated commercial baby-powder lead value and a single baby-oil dataset are held out of the summary for stated basis and dispersion reasons rather than dropped. A thirty-product talcum set is held out because its digest volume is unstated and it does not separate baby from adult powders, and a fifty-six-product talcum set is adjacent because it is predominantly adult facial talc and its three-product baby subgroup is not separately transcribed. A single 1985 bottle carrying the highest baby-powder nickel and chromium figures in the corpus is excluded from every range as an approximate abstract result. Three folk and traditional powders, two of them lead-oxide preparations, are named as a hazard class rather than as commercial occurrence data, and the only release fractions in this file belong to one of them. Market detection rates cannot be calculated from a sixteen-product convenience sample, a four-product sample, a single regulatory re-test, a single lotion, or a news-reported powder screen. The two multi-product talcum sets disagree by roughly three orders of magnitude on cadmium, and nothing here resolves that.

The research agenda follows those gaps, in order. First, any properly reported baby-oil measurement, in milligrams per kilogram as sold, with recovery and detection limits. Second, more than one baby lotion, measured the same way. Third, talc-versus-starch powder surveys that print limits, reference materials, and recoveries. Fourth, speciation of arsenic, mercury, and chromium, and the addition of tin and aluminium to the panel. Fifth, release into a sweat simulant, reported per area and time, for nickel and hexavalent chromium. Sixth, a dermal-absorption experiment on intact and compromised infant skin, or an explicit statement that the fraction is unknown.

Peer review state

This synthesis claim has not yet been evaluated by external reviewers. Verdicts will be added here as named domain experts, listed at Curators and conflict-of-interest disclosure, complete their review. Reviewer verdicts are recorded on this page.

The Heavy Metal Index publishes synthesis claims as preprints — before external review completes — with the review state visibly tracked. Until at least one external verdict is recorded below, a synthesis page is a preprint, not a peer-reviewed work. External review accumulates over time, and the credibility of the claim is partly the cumulative result of that visible review.

How to cite

Pendergrass, K. (2026). Heavy metals in baby lotions, oils, and powders. Institute of Contaminant Standards. https://doi.org/10.83376/hmi.heavy-metals-in-baby-lotions-oils-and-powders

References

1. Evaluated daily intake and health risk assessment of some toxic heavy metals in baby powder for various brands marketed in Benghazi, LibyaYaghi, M. M., Alhursh, Z. A., & EL Shawish, A. A. · Libyan Journal of Public Health Practices 2(2):27-46. Issue dated December 2025 on the PDF; Crossref publication date 2026-01-25 · 2026 · doi.org/10.37376/ljphp.v2i2.7556Peer-reviewed
2. The presence of NORMs and toxic heavy metals in talcum baby powderAlmugren, K. S., Abdul Sani, S. F., Muhamad Azim, M. K., Ismail, N. N., Khandaker, M. U., Alsufyani, S. J., Alkallas, F. H., Almajid, H. F., & Bradley, D. A. · Journal of Radiation Research and Applied Sciences 16:100660 · 2023 · doi.org/10.1016/j.jrras.2023.100660Peer-reviewed
3. NMPA notice of 27 non-compliant cosmetic batches (news report)Luo, N. · China Food and Drug Network (cnpharm.com), 31 December 2019, reporting the National Medical Products Administration notice of 26 December 2019 · 2019 · m.cnpharm.comNews
4. Evaluation of heavy metals in cosmetic products and their health risk assessmentArshad, H., Mehmood, M. Z., Shah, M. H., & Abbasi, A. M. · Saudi Pharmaceutical Journal 28(7):779-790 · 2020 · doi.org/10.1016/j.jsps.2020.05.006Peer-reviewed
5. Test of 20 infant powders by the Shenzhen Consumer Council (news report)Yang, L., & Tan, A. · Southern Metropolis Daily, 19 June 2020, reporting a Shenzhen Consumer Council comparative test · 2020 · m.mp.oeeee.comNews
6. Heavy metal content of selected personal care products (PCPs) available in Ibadan, Nigeria and their toxic effectsOmenka, S. S., & Adeyi, A. A. · Toxicology Reports 3:628-635 · 2016 · doi.org/10.1016/j.toxrep.2016.07.006Peer-reviewed
7. Quality and safety investigation of commonly used topical cosmetic preparationsAlmukainzi, M., Alotaibi, L., Abdulwahab, A., Albukhary, N., & El Mahdy, A. M. · Scientific Reports 12:18299 · 2022 · doi.org/10.1038/s41598-022-21771-7Peer-reviewed
8. Metals in children’s and consumer products and packagingStone, A. · Washington State Department of Ecology, Publication 14-04-014 (Revised June 2021) · 2021 · apps.ecology.wa.govGovernment report
9. Baby lotion tested for lead, cadmium, mercury and arsenic (community-funded laboratory testing, single lot)Rubin, T., & Lead Safe Mama, LLC · tamararubin.com blog post, posted 19 May 2025, updated 16 August 2025 · 2025 · tamararubin.comGrey literature
10. Determination of heavy metal (Cu, Pb, Ni) in talcum powder by using atomic absorption spectrometryRosnan, F. S. · Undergraduate report, Universiti Teknologi MARA · 2017Thesis
11. Determination of toxic heavy metals in different brands of talcum powderRehman, G., Bukhari, I. H., Riaz, M., Rasool, N., Khalid, A., Sattar, U., & Manzoor, H. S. · International Journal of Applied and Natural Sciences 2(2):45-52. No DOI assigned · 2013 · paper.researchbib.comPeer-reviewed
12. Nickel and chromium content in a bottle of baby powder from 1985 designated for hospital use opens questions regarding product source and contaminationKrekeler, M., Wudke, H., Murchland, M., Brown, K., Gokey, K., & Gillis, M. · Geological Society of America Abstracts with Programs 54(4), Joint North-Central and Southeastern Section Meeting, Paper 26-6 · 2022 · doi.org/10.1130/abs/2022NC-374622Conference abstract
13. Trace metals in cosmetic facial talcum powders marketed in NigeriaNnorom, I. C. · Toxicological and Environmental Chemistry · 2011 · doi.org/10.1080/02772248.2011.577075Peer-reviewed
14. Lead poisoning in a 6-month-old infant: a case reportDuan, Y., Yan, L., Gao, Z., & Gou, Y. · Frontiers in Public Health 11:1132199 · 2023 · doi.org/10.3389/fpubh.2023.1132199Peer-reviewed
15. Toxicants in folk remedies: implications of elevated blood lead in an American-born infant due to imported diaper powderKarwowski, M. P., Morman, S. A., Plumlee, G. S., Law, T., Kellogg, M., & Woolf, A. D. · Environmental Geochemistry and Health 39(5):1133-1143 · 2016, print issue October 2017 · doi.org/10.1007/s10653-016-9881-6Peer-reviewed
16. Childhood lead poisoning associated with traditional Chinese medicine: a case report and the subsequent lead source inquiryLin, G. Z., Wu, F., Yan, C. H., Li, K., & Liu, X. Y. · Clinica Chimica Acta 413:1156-1159 · 2012 · doi.org/10.1016/j.cca.2012.03.010Peer-reviewed
17. Guidance on heavy metal impurities in cosmeticsHealth Canada · effective 2012-07-20, modified 2017-10-02 · 2012 · canada.caGuidance
18. Quality of natural health products guideHealth Canada, Natural and Non-prescription Health Products Directorate · version 3.1, 1 May 2015 · 2015 · canada.caRegulatory text
19. Diaper rash products monographHealth Canada, Natural and Non-prescription Health Products Directorate · monograph dated 2024-02-23 · 2024 · webprod.hc-sc.gc.caRegulatory text
20. Technically avoidable heavy metal contents in cosmetic productsGerman Federal Office of Consumer Protection and Food Safety (BVL) · Journal of Consumer Protection and Food Safety 12(1):51-53 · 2017, online 6 October 2016 · doi.org/10.1007/s00003-016-1044-2Peer-reviewed
21. Washington State Toxic-Free Cosmetics Act (HB 1047), Chapter 70A.560 RCWWashington State Legislature · Chapter 455, Laws of 2023, codified as Chapter 70A.560 RCW · 2023 · app.leg.wa.govRegulation
22. 21 CFR 73.1550 Talc (color additive exempt from certification, for drugs)US Food and Drug Administration · Code of Federal Regulations, Title 21, 2025 annual edition · 2025 · govinfo.govRegulation
23. Talc (INS No. 553(iii)) specifications, prepared at the 61st JECFA (2003)Joint FAO/WHO Expert Committee on Food Additives · FAO Food and Nutrition Paper 52 Add 11 · 2003 · fao.orgRegulation

Update history

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