How do dermal absorption percentages from 3D skin models compare with in vivo human dermal studies for PFOA and short‑chain PFAS?
This fact-check may be outdated. Consider refreshing it to get the most current information.
Executive summary
In vitro 3D human skin equivalent models report substantially higher dermal uptake fractions for many PFAS—including a PFOA permeation of ~13.5% with ~38% retained in skin and much larger fractions for some short‑chain PFAS—than conservative human in vivo-based estimates, which place PFOA fractional absorption near ~1.6% under realistic product matrices [1] [2] [3]. The divergence reflects real differences in materials and methods (vehicle, dose condition, exposure duration, retention versus systemic appearance) and the very small human in vivo evidence base, so direct numerical parity cannot be assumed [4] [3] [5].
1. Why the 3D skin models show higher percentages — a clear signal for short‑chain PFAS
The Environment International study using 3D human skin equivalents exposed to PFAS in a methanol vehicle found 15 of 17 PFAS had ≥5% absorption, with PFOA permeation of 13.5% and 38% of the applied dose retained in the skin; shorter‑chain chemicals (for example, perfluoropentanoic acid) showed far higher absorbed fractions—reported as ~59% for PFPeA in the same work—while uptake generally decreased with increasing carbon chain length [4] [1] [6].
2. What human in vivo and ex vivo studies report — much lower systemic uptake but persistent uncertainty
A critical review of the human data concluded only four direct human studies measured PFOA dermal kinetics and recommended provisional fractional absorption for PFOA of about 1.6% under finite‑dose, realistic product conditions, with corresponding low steady‑state flux and Kp values—orders of magnitude lower than some in vitro-derived permeation values—underscoring that human in vivo evidence points to limited systemic uptake compared with some in vitro findings [3]. Ex‑vivo human skin Franz cell experiments and other lab studies show measurable fluxes for both short and long chain PFAS and report permeability coefficients of similar magnitude across species and chains (for example, fluxes for PFOA, PFHxA, PFBA reported in the 10^-6 range and Kp in the 10^-8 cm/h range), which supports the idea that skin can be permissive under certain conditions but does not settle systemic uptake in living humans [5] [7].
3. Why numbers differ: vehicle, dose regime, retention vs systemic measurement
Key methodological differences explain much of the gap: the 3D model experiments often used an organic solvent vehicle (methanol) and finite but high applied doses with exposure windows of 24–36 hours and then quantified fraction permeated versus fraction in the tissue, whereas in vivo human studies measure systemic appearance in blood after application in realistic product matrices (sunscreen or aqueous formulations) and under normal skin conditions; these divergent matrices and endpoints inflate apparent in vitro permeation relative to what appears in the circulation in vivo [4] [1] [3] [8].
4. Short‑chain PFAS: potentially higher dermal bioaccessibility but limited human confirmation
Both the 3D skin equivalent data and several in vitro human skin studies highlight that many short‑chain PFAS can permeate skin more readily than longer‑chain analogues—shorter chain length correlates with higher absorbed fraction in the 3D models and some ex‑vivo systems have identified short‑chain compounds (e.g., certain butane‑sulfonamide or C5 acids) with measurable permeation—yet there are few controlled human in vivo experiments to confirm how much of that material ultimately reaches blood or contributes to body burden [4] [9] [5].
5. Bottom line for interpretation and risk use
The best characterization today is that 3D human skin equivalents reveal a higher potential for dermal uptake—notably for short‑chain PFAS and under solvent‑enhanced conditions—while conservative human in vivo estimates for PFOA systemic absorption are much lower (≈1.6%) and human evidence remains sparse and variable; therefore 3D model percentages should be read as an important hazard signal and upper‑bound permissivity under test conditions, not an automatic prediction of systemic exposure in real‑world product use without adjustment for vehicle, dose, exposure frequency, and skin retention dynamics [1] [3] [4].