Exposure Assessment to Dioxins from the Use of Tampons and Diapers

Overview

DeVito and Schecter performed a screening-level and refined exposure assessment of dioxin (PCDD/PCDF) uptake from tampons and disposable diapers, motivated by late-1990s public concern that chlorine-bleached pulp products in feminine-hygiene and infant-care sanitary items might be a meaningful dioxin exposure pathway. The analytes are the 17 2,3,7,8-chlorine-substituted polychlorinated dibenzo-p-dioxin and dibenzofuran congeners; no heavy metals were measured. This page is therefore ingested as out-of-core-scope tampon and diaper contaminant context, parallel to [[sources/archer2005-tampon-dioxins-furans]] for tampons and the ANSES/BfR diaper-safety opinions for diapers. It carries the agency-level null finding for HMI’s core analytes in these product classes (the most potent dioxin congeners TCDD and 1,2,3,7,8-pentachlorodibenzo-p-dioxin were not detected in any sample) and the population-frame exposure parameters that may be reusable for HMI’s future heavy-metals dermal-exposure modelling in diapers and menstrual products.

Key numbers

Analytical panel and detection limits (p. 25)

• Analytes: the 17 2,3,7,8-chlorine-substituted polychlorinated dibenzo-p-dioxin and polychlorinated dibenzofuran congeners included in the WHO toxic-equivalency-factor (TEF) methodology.
• Detection limits: 0.1-0.2 ppt (pg/g product) for the individual chemicals (Table 1 footnote).
• Laboratory: ERGO Research Company, Hamburg, Germany (WHO-certified for dioxin analysis); the extraction/analysis followed the laboratory’s previously-published method (refs 1, 2, 25).
• Reference materials and instruments: not reported in the body of the paper; the method citation is to ERGO’s prior PCDD/PCDF analytical work.

Total dioxin concentrations in tampons (Table 1, pg/g product)

• Brand A (rayon): 1.5
• Brand A (cotton): 3.1
• Brand B: 23.6
• Brand C: 8.7
• Brand D: 7.7
• Octachlorodibenzo-p-dioxin (OCDD) accounted for 48-88% of the total mass of dioxins in the tampons; the remaining congeners made up <1-16% of the total mass.
• TCDD (2,3,7,8-tetrachlorodibenzo-p-dioxin, the most potent congener) and 1,2,3,7,8-pentachlorodibenzo-p-dioxin were not detected in any sample.

Total dioxin concentrations in diapers (Table 1, pg/g product)

• Brand E disposable: 3.7
• Brand F disposable: 2.2
• Brand G disposable: 1.8
• Conventional cotton diaper: 2.5
• Only 5 of the 17 congeners were detected in the diapers.
• OCDD accounted for 67-76% of the total dioxins in the diapers. Octachlorodibenzofuran (OCDF) and the heptachlorinated dibenzo-p-dioxins together accounted for 5-21%. 2,3,7,8-TCDF accounted for 2-6% of the total.

TEQ concentrations (Table 2, pg TEQ/g product; calculated using WHO TEFs)

• Tampons: Brand A rayon 0.013, Brand A cotton 0.015, Brand B 0.034, Brand C 0.020, Brand D 0.24. Brand D is a ~10-fold outlier on TEQ relative to the other brands despite having a similar total-dioxin mass to Brand C, because Brand D was the only sample with detectable 2,3,4,7,8-pentachlorodibenzofuran (one of the highest-TEF congeners; TEF = 0.5).
• Diapers: Brand E 0.023, Brand F 0.013, Brand G 0.013, conventional cotton 0.0042. The text on p. 26 states the range as 1.6-3.0 pg TEQ/g diaper while citing Table 1; this conflates Table 1 (total dioxin mass concentrations, 1.8-3.7 pg/g) with Table 2 (TEQ concentrations, 0.0042-0.023 pg TEQ/g). The tables themselves are internally consistent; the verbal range is a copy-edit slip that this page does not propagate.
• TCDF concentration is at the detection limit (0.1 ppt) in tampon brands A, B, C and in two of the three disposable diapers; the authors caution these values should be viewed with suspicion. TCDF was not detected in the conventional cotton diaper, which accounts for most of the TEQ difference between the cotton and the disposable diapers.
• The hexachlorinated dibenzo-p-dioxins, TCDF, and 2,3,4,7,8-pentachlorodibenzofuran together account for >90% of the TEQs in Brand D tampons. TCDF and the heptachlorinated dibenzo-p-dioxins together account for ~90% of the TEQs in the lower-TEQ tampon brands.

Exposure-model assumptions (p. 24)

• Tampon-use scenario: 6 tampons/day for 5 days/month, averaged over 30 days. Adult female body weight 60 kg. Average tampon mass per brand: A rayon 4.73 g, A cotton 3.15 g, B 1.9 g, C 3.4 g, D 4.04 g (Table 3).
• Diaper-use scenarios: infants 0-6 months use 10 diapers/day, average body weight 6.75 kg; toddlers 6-24 months use 6 diapers/day, average body weight 11 kg. Average diaper mass 40 g. Urine load 45 g/diaper (refined analysis).
• Partition coefficient (K_p): 5340 for 2,3,7,8-TCDF from pulp to synthetic urine over 8 hr, used as a deliberately-conservative proxy for all dioxin congeners (TCDF is the most water-soluble, so applying its K_p to higher-chlorinated congeners overestimates their bioavailability).
• Menstrual-fluid volume: 10 g/tampon (manufacturer package-insert estimate).
• Dermal absorption fraction: 3% for the screening-level diaper analysis (U.S. EPA upper-bound estimate from low-organic-content soil and from PCDD/F transfer to skin from polyester/cotton fabrics over 72 hr, refs 31, 32); 28% for the refined diaper analysis (in-solution dermal absorption from aqueous vehicles, ref 8).

Estimated daily intakes (screening-level analysis; Tables 3 and 4)

• Tampons (adult, pg TEQ/kg/day): A cotton 0.00083, A rayon 0.0014, B 0.0013, C 0.0013, D 0.019. Dietary adult intake assumed 1 pg TEQ/kg/day (U.S. EPA, ref 8). Diet/tampon ratios range from 51 (Brand D) to 1211 (Brand A cotton); Brand D accounts for ~3.3% of the adult daily total-TEQ exposure under this analysis.
• Diapers, nursing infant 0-6 mo (pg TEQ/kg/day): E 0.041, F 0.023, G 0.023, cotton 0.0075. Breast-milk intake assumed 145 pg TEQ/kg/day (= 980 pg TEQ/day / 6.75 kg; ref 8). Diet/diaper ratios 3498-19,374.
• Diapers, toddler 6-24 mo (pg TEQ/kg/day): E 0.013, F 0.0072, G 0.0072, cotton 0.0023. Dietary intake assumed 3.6 pg TEQ/kg/day (= 40 pg TEQ/day / 11 kg; ref 8). Diet/diaper ratios 283-1568.

Estimated daily intakes (refined analysis with bioavailability partitioning; Tables 3 and 4)

• Tampons: approximately 100-250 times lower than the screening-level estimates. Diet/tampon ratios 1.1 x 10^4 (Brand D) to 2.0 x 10^5 (Brand A cotton); equivalently dietary exposure is 13,000-240,000 times the tampon exposure.
• Diapers, 0-6 mo: diet/diaper ratios 4.0 x 10^5 (Brand E) to 2.2 x 10^6 (cotton).
• Diapers, 6-24 mo: diet/diaper ratios 3.2 x 10^4 (Brand E) to 1.8 x 10^5 (cotton).
• Overall: dietary dioxin exposure is 30,000-2,200,000 times the dermal-uptake exposure from diapers under the refined analysis.

Authors’ overall conclusion

The authors conclude that exposure to dioxins through normal use of tampons and infant diapers does not significantly contribute to U.S. dioxin exposure, and that the trace-level concentration differences between cotton and pulp-based products are small enough that there is no significant difference in dioxin exposure between the two material classes. They caution that several of the detected congeners — particularly TCDF and the heptachlorinated dibenzo-p-dioxins — are at the analytical detection limit and should be viewed cautiously; substituting one-half the detection limit for non-detects (a common convention they explicitly declined) would raise the TEQs by approximately one order of magnitude but would still leave tampon and diaper exposures at <1% of the dietary TEQ.

Methods (brief)

Sampling: tampons and diapers purchased in 1997 in San Francisco, California, by volunteers from Mothers and Others for a Livable Planet. All boxes of a given product were from the same lot number. Tampon brands A and B and disposable diaper brands E and F and the cotton diaper were sourced from the same large department store; Brand A tampons were available in rayon and cotton variants and both were analyzed separately. Brand C tampons and Brand G disposable diapers were sourced from a health-food-store chain. Brand D tampons were ordered from a specialty company. Products were shipped in their original commercial containers to ERGO Research Company (Hamburg, Germany).

Analysis: PCDD/PCDF congeners were extracted and quantified at ERGO Research Company per their previously-published method (refs 1, 2, 25). ERGO is certified by the World Health Organization for dioxin analysis. The 17 2,3,7,8-chlorine-substituted PCDD/PCDF congeners in the WHO TEF methodology were targeted; non-detects were assigned a concentration of zero in TEQ summation (the authors note that the alternative one-half-LOD convention would have raised TEQ estimates by approximately one order of magnitude).

Exposure modelling: two analyses were performed for each product class. A screening-level analysis assumed 100% bioavailability and a 3% (diapers) or full-release (tampons) absorption fraction. A refined analysis used the 2,3,7,8-TCDF pulp-to-synthetic-urine partition coefficient (K_p = 5340, ref 27) to estimate the fraction of dioxin partitioning out of the product matrix, with 28% dermal absorption for the dissolved fraction in the diaper analysis. Dietary intake comparators came from U.S. EPA (ref 8): 1 pg TEQ/kg/day for adults, 145 pg TEQ/kg/day from breast milk for nursing infants, and 3.6 pg TEQ/kg/day for toddlers.

Implications

Microbiome: Not measured. The paper discusses endometriosis and dioxin exposure as motivation but does not measure microbiome, bacterial, or reproductive-tract endpoints.

Related evidence

• Diapers and diaper components
• Menstrual tampons (scented/unscented; applicator tested as distinct component)
• Opinion of the French Agency for Food, Environmental and Occupational Health & Safety on the safety of baby diapers (revised opinion, Request No 2017-SA-0019)

Update history

The five most recent substantive edits to this page, classified major (evidence or structure moved), correction (a published value or statement was wrong and has been fixed), or minor (narrative rewritten without changing the underlying evidence). Each description is derived from what the edit did to this page; the linked commit is the authoritative record, routine regeneration passes are excluded, and the full version history lives in git.