TL;DR — Peer-reviewed research converges on four facts about the eye area. Eyelid skin is the thinnest on the face (758.9 µm dermis, cadaver study). Less rubbing measurably improves skin condition and barrier function (n=35 single-arm trial). Eyeliner reaches the tear film within 5–10 minutes when it is applied inside the lash line. That comes from a 3-subject pilot, backed by larger cross-sectional studies with a 2026 counterweight. Used makeup sponges culture at over 10^6 CFU/ml. And 79–90% of all donated products carried 10^2–10^3 CFU/ml. The evidence confirms conventional advice. Stay outside the lash line, clean the tools, and keep friction low. It also gives the criteria behind this series' tool comparison and cleaning guide.
Upper-eyelid skin is the thinnest on the human face. Its dermis measured 758.9 micrometers at the thinnest facial subunit in a cadaveric biopsy study. That one number anchors the rest of what the peer-reviewed record says about makeup and the eye area. Rubbing measurably worsens skin condition. Eyeliner applied inside the lash line reaches the tear film within 5 to 10 minutes. And the worst-kept makeup sponges have been cultured at more than one million bacteria per milliliter. This article reads that record study by study. Each study's method, sample size, and limits are set out next to its finding. It does not evaluate products. It does not give medical advice.
Figure 1 — Four variables, four evidence bases
The thinnest skin on the face, by actual measurement
The 758.9-micrometer figure comes from a 2015 study in Aesthetic Surgery Journal. The team took full-thickness skin biopsies from 39 defined facial subunits. The tissue came from 10 cadaveric heads. The team then measured dermal and total skin thickness histologically. The upper medial eyelid had the thinnest dermis of any facial region measured. It also had the least total skin thickness of the regions measured.
The caveats are real. This is a cadaveric study, and the sample is ten heads. Formalin fixation can make the tissue shrink. So it is best to read the absolute micrometer values as a ranking, not as a precise in-vivo measurement. The abstract does not report any demographic breakdown. What survives the caveats is the ranking itself. Whatever the exact number, the skin around the eye has less mechanical margin. That holds against any other place a person routinely applies and removes pigment.
Friction is a measured variable, not a metaphor
The clearest evidence that force damages skin, and not just chemistry, is a 2021 study in the Journal of Cosmetic Dermatology. Thirty-five women with mild atopic dermatitis took part. They were switched to a cleansing oil with higher cleansing ability. It took less rubbing force to get the makeup off. Over four weeks, dryness, scaling, irritation, erythema, and itchiness all decreased significantly, and transepidermal water loss fell as well. Transepidermal water loss is an objective measure of the skin barrier. The authors read it as compensation: the women had been scrubbing harder to make up for weak cleansers, and worsening their skin without knowing it.
The design limits what the study can prove. It is a single-arm intervention with no control group. So the drop in rubbing cannot be fully separated from the changed formulation. The women also had atopic dermatitis rather than average skin. But the finding points the same way as uncontroversial expert advice. The American Academy of Dermatology tells consumers to cleanse with the fingertips and to resist scrubbing, because scrubbing irritates the skin. The same page says not to pull hard on the delicate skin around the eyes. That page is patient education, not graded evidence. But it is consistent with the measured result.
The lid margin: mechanism first, then the studies
To see why placement matters, start with how the eye normally works. The tear film's outer lipid layer is built from meibum. Meibum is an oil secreted by the meibomian glands along the eyelid margin. Each blink spreads that oil up over the surface of the eye. That comes from the TFOS DEWS II report. It is the ophthalmology field's consensus reference on tear film physiology. This is mechanism, not a claim about cosmetics. But it explains the rest. Anything left at the lid margin sits right in the path of the eye's own lubrication system.
The migration evidence itself starts very small. A 2015 pilot study in Eye & Contact Lens applied a glitter-containing pencil eyeliner to three subjects. It tracked the particles by slit lamp. Applied inside the lash line, the product reached maximum tear film contamination within 5 to 10 minutes. Application outside the lash line produced slower and reduced migration. Three subjects, one product, visual tracking rather than chemical assay. This is a proof of mechanism, not a population estimate.
Larger observational studies point the same way. A 2020 cross-sectional study in Cornea compared 42 Thai women aged 18–40. It found tear breakup time of 3.0 ± 1.9 seconds in regular eyeliner users versus 5.8 ± 2.1 seconds in non-users (P < 0.001). The same study graded the meibomian glands as worse in those users. That is an association, not proof of causation. The paper's title sounds more causal than that. A 2022 study in the Saudi Journal of Ophthalmology took the pattern further. It covered 220 healthy women. Users of eyeliner, mascara, or both showed significantly greater meibomian gland loss on meibography and shorter non-invasive tear breakup time than non-users. Schirmer testing found no difference in tear production. The study is cross-sectional and single-center. So the same causation caveat applies. But the null Schirmer result is itself informative. Whatever is happening involves the oil glands, not the tear-producing ones.
Then comes the counterweight. It belongs in any honest reading. A 2026 study in the Journal of Clinical Medicine compared three groups with a masked examiner. The groups were 10 non-users and 18 women applying only outside the eyelid margin. A further 21 applied directly on the margin. It found worse gland secretion quality and more ocular surface staining in makeup users. But there was no between-group difference in meibomian gland dropout, lipid layer thickness, or tear breakup time. The groups are small. Still, the study partly contradicts the louder findings above. The structural gland-loss signal did not replicate here. The fairest summary of the whole set is a split one. Eyeliner at the lid margin is consistently associated with functional tear film measures. The structural findings vary between studies.
Expert guidance already reflects these studies. The American Academy of Ophthalmology advises applying eye makeup outside the lash line, away from the eye, to avoid blocking the eyelid's oil glands. That is a consumer education page rather than a graded clinical guideline. But it is the field's own summary. It matches what the placement studies measured.
The tools carry their own numbers
The tool is a separate variable from the product. And it has been cultured. A 2020 study in the Journal of Applied Microbiology collected used cosmetic products donated by UK consumers. It found 79–90% of them contaminated with bacteria at 10^2 to 10^3 CFU per milliliter, with makeup sponges averaging more than 10^6 CFU per milliliter. Staphylococcus aureus and E. coli were among the organisms detected. Of the sponges, 93% had never been cleaned. Also, 64% had been dropped on the floor and used again. The design caveats matter. Donated products self-select for owners willing to part with old makeup. The sample is modest. Colony counts are not illness counts. The study shows contamination, not documented infections.
But the key detail is which tool carried the worst load. The sponge is a reusable product. It is porous, damp-stored, and, in this sample, almost never washed. The contamination shown here is not a property of reuse itself. It is a property of reuse without maintenance. It is concentrated in materials that are hard to clean. That is why expert guidance attaches a schedule to reuse. The American Academy of Dermatology tells consumers to wash makeup brushes every 7 to 10 days and never to share them. That is an expert-consensus interval, not a trial result. But it is a schedule nonetheless.
What the record supports — and what it settles
Read together, the studies confirm rather than overturn the conventional advice. The eye area's skin is the face's thinnest. Mechanical rubbing is a measured aggressor. Product placed at the lid margin migrates into the tear film. An unmaintained tool is a microbial reservoir. Nothing in this record shows that makeup wearers are ruining their eyes. The 2026 counterweight study is the reminder of two things. The structural findings are contested, and the samples are small. What the record does settle is which variables matter: surface thinness, friction, placement, and tool hygiene.
Those four variables are exactly the axes on which the tools differ. Consider a bare fingertip, a disposable cotton bud, a brush, a sponge, or a reusable precision swab with a thermoplastic elastomer (TPE) tip. Each sits differently on precision, applied force, and cleanability. The next articles put those variables to work. Precision Tool vs. Bare Fingertip turns this evidence base into six criteria for choosing between a precision tip and your finger. The cleaning guide follows the contamination numbers above to a maintenance method that holds up. It also explains why the alcohol shortcut fails. The research reported here is the reason those criteria exist. The comparisons are where it becomes usable.
This magazine is published by LastObject, a company that manufactures reusable personal-care tools. This article names no brands, compares no products, and reports only external, independently published research.
