tapelift-clock-2026

TapeLift Epigenetic Clock (Rodríguez-Paredes 2026)

The TapeLift clock is a non-invasive skin DNA methylation aging clock trained on epidermal cells collected by adhesive gel tape from the forehead, published in npj Precision Oncology (2026). It removes the principal barrier to epidermal methylation profiling — the requirement for punch biopsy under local anaesthesia — by demonstrating that ~49 ng of genomic DNA from four sequential adhesive strips yields sufficient input for Illumina MethylationEPIC v2 array. Two clock variants are reported: a standard elastic-net model (157 CpGs) and a principal-component (PC) version (5,021 CpGs) with better independent validation accuracy.

Critical note (R40 verifier finding): The training cohort is Caucasian-only (Fitzpatrick I–IV). The “no non-white bias” property frequently attributed to TapeLift in narrative summaries originates from Qi 2026’s independent external application of the TapeLift clock to an n=17 multi-ethnic Brazilian pilot cohort, not from any test in the original Rodríguez-Paredes 2026 paper. The biomarker page bormann-epidermis-clock-2016 and qi-23k-epidermis-clock-2026 should be consulted as R43 sisters for the cross-ethnic validation provenance.

Identity and Origin

• Primary citation: Rodríguez-Paredes M et al. (2026) npj Precision Oncology — doi:10.1038/s41698-026-01302-7
• PMID: 41593290 · PMC: PMC12948955 (gold OA; CC BY 4.0)
• Archive status: local PDF available at
• Study page: rodriguez-paredes-2026-tapelift-clock (R39 verified, full-scope PDF read 2026-05-19)
• Clock family: Tissue-specific (epidermis); second-generation (chronological-age trained)
• Input modality: Illumina MethylationEPIC v2 array; tape-strip epidermal keratinocytes
• Affiliation: DKFZ (German Cancer Research Center) + Beiersdorf AG collaboration (Lyko/Grönniger group)
• Framing context: Paper is primarily a skin cancer biomarker study; the aging clock is one component within a broader skin cancer epigenomics framework. This distinguishes it from pure aging-clock papers like horvath-clock-2013 or dunedinpace-2022.

Clock Performance

Two clock variants were reported ¹:

Clock	CpGs	Training RMSE	Internal validation RMSE (n=18)	Independent validation RMSE (n=19)
Elastic-net (α=0.1)	157	6.0 yr	5.8 yr	6.2 yr
PC-based (Higgins-Chen protocol)	5,021	4.7 yr	4.3 yr	5.1 yr

The PC clock outperforms the elastic-net clock at every evaluation stage. The improvement follows the Higgins-Chen 2022 observation ² that substituting principal-component loadings for raw CpG β-values reduces technical batch-effect noise, improving both reliability and validation accuracy. The independent validation RMSE of 5.1 yr (PC clock) is comparable to the Bormann 2016 epidermis clock (MAE ~4–5 yr cross-validated) and slightly worse than the Menendez Vazquez 2025 non-invasive skin clocks (MitraSolo MAE 4.09 yr / RMSE 5.25; MitraCluster MAE 4.00 yr / RMSE 5.23; both on independent set n=75, R² 0.88–0.89) ³.

Test-retest ICC: Not reported in the original paper. long-term-unknown — ICC, longitudinal stability, and diurnal variation are not characterized in Rodríguez-Paredes 2026.

Methodology: TapeLift Sampling Protocol

Procedure: Four sequential adhesive strips of polyacrylate-coated polyurethane film are applied to a 2.0 × 2.5 cm² delineated area on the forehead (pre-cleaned with 70% ethanol), capturing upper stratum corneum and outer viable epidermis (keratinocytes). Mean DNA yield: ~49 ng per session. Only 20% of extracted gDNA goes to qPCR QC; 80% is reserved for microarray ¹.

Two-stage QC pipeline:

1. Pre-array DNA integrity: qPCR ΔCt <3 cutoff → 95/105 donors passed (90.5%)
2. Post-array probe detection: >75% probes with p<0.05 → 89/95 passed (93.7%)

High-confidence probe set: A four-stage filtering pipeline (remove polymorphic, non-autosomal, cross-reactive, and high-intra-individual-variance probes) yielded 535,384 high-confidence CpG probes from the ~833,305 EPIC v2 probes, used for all downstream clock training and cancer analyses.

This contrasts with the Bormann 2016 approach ⁴, which required invasive punch biopsy (3 mm) under local anaesthesia. TapeLift is procedurally equivalent to a standard cosmetics consumer-research protocol, enabling broader deployment in epidemiological and product-efficacy studies.

Tumor-Suppressor Gene Angle

A distinctive feature of the TapeLift study is its explicit integration of epigenetic aging with skin cancer biology. Starting from the TSGene 2.0 database (1,217 TSGs), the authors identified 859 TSGs with adequate promoter CpG coverage in the high-confidence probe set. The 50 TSGs showing the most pronounced age-associated promoter hypermethylation (p<0.05, Pearson r with age) were then evaluated in actinic keratosis (AK) and cutaneous squamous cell carcinoma (cSCC) samples ¹.

Named examples with effect sizes:

• ZIC1 (r=0.6, p=2.7×10⁻¹⁴)
• RBM38 (r=0.5, p=1.3×10⁻⁴)
• FOXA2 (r=0.5, p=5.9×10⁻⁷)
• UCHL1 (r=0.4, p=6.7×10⁻⁴)
• ID4

Mean promoter methylation across these 50 TSGs was further elevated in AK and cSCC compared to age-matched normal epidermis, suggesting that age-associated epigenetic silencing of tumor suppressors represents a field cancerization mechanism for keratinocyte cancer (KC) risk. An associated mitotic age signal (epigenetic measure of cumulative cell divisions) also showed progressive elevation from normal epidermis → AK → cSCC, positioning mitotic age as a potential KC risk-stratification biomarker.

This TSG-aging angle connects epigenetic-alterations to cancer in a tissue-specific context not well-covered by pan-tissue clocks. no-mechanism — whether epigenetic TSG silencing is causally upstream of cancer initiation or correlates with cancer-prone tissue biology is not established.

Cross-Ethnic Validation Framing (R40-Verified Critical Correction)

What the original paper shows: Caucasian-only training (Fitzpatrick I–IV). No non-white donors in training or independent validation cohorts. Cross-ethnic performance not tested.

Where the “no non-white bias” claim comes from: Qi 2026 ⁵ applied the TapeLift clock as a comparator clock to their independent n=17 multi-ethnic Brazilian pilot cohort (Fitzpatrick I–VI) and reported no detectable ethnic bias for either the 23k clock or TapeLift. This external evaluation by a different research group (same Beiersdorf/DKFZ affiliation, but independent cohort) is the sole source of the cross-ethnic claim.

Implications: Narrative reviews and wiki pages that attribute cross-ethnic robustness to TapeLift as an intrinsic property of the Rodríguez-Paredes 2026 paper are conflating the primary paper with the Qi 2026 external validation. The n=17 cross-ethnic pilot is underpowered to definitively establish absence of ethnic bias; a properly powered, ethnically diverse replication cohort is needed. needs-replication

Funding and Competing Interests

Per R40 verifier finding:

• Funding: Open Access publication fee funded via Projekt Deal (German university consortium OA agreement with Springer Nature). No Beiersdorf research grant is credited for the study.
• Competing interests: K.W. and E.G. are Beiersdorf AG employees; F.L. received consultation fees from Beiersdorf AG; M.R.-P. received speaker fees from Beiersdorf SA. All four named authors (including M.R.-P., K.W., E.G., F.L.) are listed as inventors on patent applications related to the TapeLift method.
• License: CC BY 4.0 (not CC BY-NC 4.0; confirmed via PDF and PMC record).

The Beiersdorf commercial interest is in the TapeLift sampling method (patent applications filed), not in a specific consumer product outcome. This does not invalidate the clock training results but provides context for independent replication priority.

Comparison to Related Clocks

Clock	Year	Tissue	Sampling	CpGs	Validation RMSE/MAE	Ethnicity	Key context
Bormann 2016 4	2016	Epidermis	Punch biopsy (invasive)	~500	MAE ~4–5 yr (cross-val)	White females only	Foundation for Qi 2026 23k clock
TapeLift elastic-net (this page)	2026	Epidermis	Tape strip (non-invasive)	157	RMSE 6.2 yr (indep test)	Caucasian only	Skin cancer context
TapeLift PC clock (this page)	2026	Epidermis	Tape strip (non-invasive)	5,021	RMSE 5.1 yr (indep test)	Caucasian only	Better accuracy
Qi 23k clock qi-23k-epidermis-clock-2026	2026	Epidermis	Tape strip (non-invasive)	~23k	Cross-val MAE 5.66 yr; multi-ethnic 4.88 yr	Multi-ethnic incl. Fitzpatrick V–VI	DHM cosmetic study context
Menendez Vazquez MitraSolo/MitraCluster 3	2025	Epidermis	Tape strip (non-invasive)	MitraSolo 3,831 CpGs / MitraCluster 931 regions (14,089 CpGs)	MitraSolo MAE 4.09 / RMSE 5.25; MitraCluster MAE 4.00 / RMSE 5.23 (n=75 independent)	Multi-ethnic, all Fitzpatrick types; UK + US sites	Largest training set (n=387 train / 75 test from 590 individuals); PCA-based ElasticNet on enzymatic methyl-seq
Horvath 2013 horvath-clock-2013	2013	Pan-tissue (blood / 51 tissues)	Blood draw / biopsy	353	MAE ~3.6 yr	Diverse	Standard reference clock
PhenoAge phenoage-2018	2018	Blood	Blood draw	513	n/a (mortality-trained)	Diverse	Mortality-trained

The Menendez Vazquez 2025 clocks ³ are a notable recent competitor: they use enzymatic methyl-sequencing (NEBNext EM-seq on Illumina NovaSeq) rather than microarray, are trained on a 590-individual database (n=387 training / n=75 independent test from an n=462 development subset; 817-sample total), use a PCA-based ElasticNet architecture, achieve MAE ~4.0–4.1 yr (R² 0.88–0.89) on the independent set, and generalize across anatomical sites, sampling methodologies, and Fitzpatrick skin types. Intra-individual variation over 9 months: 1.76 yr (MitraSolo) / 1.67 yr (MitraCluster). These clocks were not available when Rodríguez-Paredes 2026 was developed but are now the primary comparators for non-invasive skin methylation clocks. needs-replication — direct head-to-head comparison of TapeLift vs MitraSolo/MitraCluster on the same cohort is not yet published.

Intervention-Responsive Status

intervention-responsive: partial — The TapeLift clock is used as a comparator in Qi 2026’s pilot study of dihydromyricetin (DHM) topical serum, where the TapeLift-defined epigenetic skin age showed reduction of ~2.1 yr at 8 weeks in the DHM cohort ⁵. However, this is from an uncontrolled open-label n=60 cosmetic study without vehicle comparison, and the result is attributed specifically to Qi 2026’s study design rather than an intrinsic property of the TapeLift clock. No dedicated intervention RCT using TapeLift as a primary endpoint has been published as of 2026-05-19. needs-replication

Limitations and Gaps

1. Caucasian-only training cohort — Fitzpatrick I–IV only; no cross-ethnic validation in the original paper. The n=17 Qi 2026 external pilot is underpowered to establish ethnic generalizability.

2. Forehead-only sampling site — No other anatomical sites validated in the original paper; Menendez Vazquez 2025 showed generalization across sites, but that is a different clock with different training. Whether TapeLift accuracy holds at sun-protected vs sun-exposed sites is unknown. long-term-unknown

3. Single-paper origin — No independent replication of the Rodríguez-Paredes clock from outside the Lyko/Grönniger/Beiersdorf group. needs-replication

4. No published test-retest ICC — Reliability across repeated sampling sessions is not characterized. By comparison, DunedinPACE ICC = 0.96 and conventional pan-tissue clocks have been benchmarked extensively. long-term-unknown

5. Cross-sectional design — Aging effects inferred from age-stratified cross-sectional samples; no longitudinal tracking of individual methylation change in the same donors.

6. Cancer-aging collinearity — The clock and the TSG hypermethylation findings were developed in the same 89-sample cohort; whether these signals are mechanistically linked or merely co-trained on the same data is not resolved. no-mechanism

7. Framing as a cancer biomarker — The paper’s primary framing is skin cancer risk / field cancerization, not aging per se. Whether TapeLift clock acceleration predicts functional aging outcomes (frailty, functional decline, mortality) is unknown. needs-replication

Cross-References

• rodriguez-paredes-2026-tapelift-clock — primary study page (R39 verified; all quantitative claims sourced from here)
• bormann-epidermis-clock-2016 — R43 sister; first epidermis methylation clock; Bormann 2016 punch-biopsy approach (implicit stub — no biomarker page yet)
• qi-23k-epidermis-clock-2026 — R43 sister; parallel 2026 tape-strip skin clock; provides cross-ethnic external validation of TapeLift (implicit stub — no biomarker page yet)
• qi-2026-dhm-epigenetic-skin-aging — Qi 2026 study page; source of the cross-ethnic TapeLift validation claim and DHM intervention data
• horvath-clock-2013 — pan-tissue reference clock
• phenoage-2018 — mortality-trained clock; blood-based comparator
• epigenetic-alterations — hallmark context; age-associated TSG promoter hypermethylation in skin supports this hallmark
• skin-aging — skin-specific aging phenotype page; TapeLift + Bormann clocks are canonical skin methylation biomarkers
• cancer — skin cancer angle (AK → cSCC progression tracked via mitotic age and TSG hypermethylation)
• epidermis — tissue cross-reference (implicit stub)
• keratinocytes — cell-type of origin for tape-strip samples (implicit stub)
• biological-age-measurement — cross-clock comparison MOC; TapeLift occupies the tissue-specific epidermis-only quadrant

Footnotes

Footnotes

1. rodriguez-paredes-2026-tapelift-clock · doi:10.1038/s41698-026-01302-7 · PMID 41593290 · PMC12948955 · n=105 enrolled (89 QC-passing; 71 train / 18 internal val; 19 independent test) · observational cross-sectional · Caucasian Fitzpatrick I–IV · elastic-net clock: 157 CpGs, indep val RMSE 6.2 yr; PC clock: 5,021 CpGs, indep val RMSE 5.1 yr · forehead tape-strip sampling · npj Precision Oncology 2026 · gold OA CC BY 4.0 · local PDF available ↩ ↩² ↩³

2. doi:10.1038/s43587-022-00248-2 · Higgins-Chen A et al. · Nature Aging · 2022 · methods/review · PC-based epigenetic clocks substantially improve test-retest reliability vs conventional elastic-net clocks; the TapeLift PC clock follows this protocol · archive: pending download (green OA via PMC9586209) ↩

3. menendez-vazquez-2025-mitrasolo-mitracluster · doi:10.1038/s41514-025-00314-0 · PMC12820032 · Menendez Vazquez A et al. (BWH/Harvard Medical School) · npj Aging 2025 · observational · n=387 train / n=75 independent test from 590-individual / 817-sample database; ages 18–90 yr; all Fitzpatrick types and ethnicities; UK + US sites · PCA-based ElasticNet on NEBNext EM-seq (Illumina NovaSeq) · MitraSolo (3,831 CpGs) MAE 4.09 / RMSE 5.25 / R² 0.88; MitraCluster (931 regions / 14,089 CpGs) MAE 4.00 / RMSE 5.23 / R² 0.89; intra-individual variation 1.76 / 1.67 yr over 9 months (n=76) · archive: downloaded (PMC OA) ↩ ↩² ↩³

4. doi:10.1111/acel.12470 · Bormann F, Rodríguez-Paredes M et al. · Aging Cell · 2016 · observational · n=108 (white females, ages 18–78) · first DNA methylation clock for human epidermis; punch biopsy (invasive); foundational training dataset for Qi 2026 23k clock · local PDF available in a local paper archive ↩ ↩²

5. qi-2026-dhm-epigenetic-skin-aging · doi:10.1007/s13555-026-01764-4 · n=77 (17 multi-ethnic pilot + 60 DHM open-label) · observational pilot + open-label single-cohort · 2026 · source of cross-ethnic TapeLift external validation (n=17, Fitzpatrick I–VI; no detectable ethnic bias reported); DHM serum reduces epidermal DNAm age ~2.1 yr at 8 wk (no vehicle control) · model: human, multi-ethnic Brazilian cohort ↩ ↩²