UV nail dryers can damage cells under laboratory conditions, but the studies discussed here do not establish how much, if at all, gel manicures increase skin cancer risk in people. That uncertainty supports reducing avoidable UVA exposure without turning a dish experiment into a personal cancer forecast. The 2024 systematic review describes the available evidence as weak and any risk from prolonged, repeated use as potentially low. Metko and colleagues, 2024.
For someone deciding whether to keep getting gel manicures, three questions matter: what the experiments actually measured, how the exposure compares with an appointment, and which precautions are reasonable while long-term human risk remains uncertain. A cell-survival result, a lamp measurement and a diagnosed cancer answer different questions. Keeping those outcomes separate makes the apparent contradictions easier to understand.
The American Academy of Dermatology recommends broad-spectrum, water-resistant SPF 30 or higher on the hands before a gel manicure, or dark, opaque gloves with the fingertips exposed. Choosing traditional polish avoids curing-lamp exposure. These are practical options; none supplies a scientifically established number of safe gel appointments. AAD gel-manicure advice.
What the 2023 cell study actually found
Zhivagui and colleagues exposed primary mouse fibroblasts, human fibroblasts and adult human epidermal keratinocytes to a 54-watt nail dryer emitting 365–395 nm light. A single continuous 20-minute exposure produced 20–30% cell death; three repeated 20-minute exposures produced 65–70%. Viability was assessed 48 hours after the final exposure. The separate acute DNA-damage protocol used two 20-minute exposures in one day; the chronic protocol used one on each of three consecutive days. These prolonged laboratory schedules differ from the shorter curing times described below. Original Nature Communications study.
The investigators also found oxidative DNA damage and mutations associated with reactive oxygen species. That establishes a biological hazard in the tested cells. It does not establish cancer incidence in clients: the experiment did not follow people, measure tumors, or reproduce intact skin with its outer protective layer. The authors explicitly cautioned against treating their results as direct evidence of increased human cancer risk. Cell-death percentages are percentages of laboratory cells, never percentages of manicure users who will develop cancer.
The correction changed wording, not the results
The publisher issued an Author Correction on March 14, 2023. It replaced an abstract sentence claiming that no experimental evaluation had been conducted with wording saying the effects on mammalian-cell physiology and mutagenesis remained unclear. The notice describes a wording correction in the abstract, not a retraction of the study or a withdrawal of its experimental results. Publisher correction notice.
This matters when assessing competing summaries. A correction does not automatically erase a finding. Equally, the continued publication of the experiment does not transform its laboratory outcome into proof of cancer in people. Read the corrected claim and the experimental design together; neither a dramatic headline nor a dismissal based solely on the word correction answers the clinical question.
Shorter exposure changes the result, but not every endpoint
Słabicka-Jakubczyk and colleagues tested a different lamp and HaCaT keratinocytes in 2023. Four minutes of exposure reduced measured viability by 8%, which was not statistically significant; 20 minutes reduced it by 35%. The paper describes typical manicure exposure as roughly 3–6 minutes, while noting that products and techniques vary. A nonsignificant finding is not proof of no effect, and this viability assay did not measure long-term cancer. Scientific Reports study.
An LED label does not mean UV-free: both fluorescent and LED curing lamps can emit UVA. Follow the matched product-and-lamp curing instructions. Lamp types and curing.
Finn and colleagues tested six 30-second exposures separated by two-minute pauses in 2024. Keratinocyte viability did not decline with that salon-style schedule, but fibroblast viability did, and DNA damage was measured even under the salon schedule. The study also found substantial heating during prolonged runs, with temperatures around 40–65°C. Heat-only and UV-shielded controls showed that temperature could harm cells independently of UVA. These results identify an important experimental confounder; they do not show that half the damage in the original study was caused by heat. International Journal of Toxicology study.
The useful interpretation is specific: the tested cells, lamp, exposure schedule and endpoint all matter. Viability describes whether cells survive an assay; surviving cells can still carry DNA damage. These follow-up studies therefore cannot support the blanket statement that realistic manicures cause no cellular injury. Nor can their dish findings quantify the cancer risk of a human hand, which handles heat and radiation differently.
Why lamp measurements cannot give you a safe visit count
Shipp and colleagues measured 17 lamps in 16 US salons in 2014. Median UVA irradiance across devices ranged from 0.6 to 15.7 mW/cm². Their modeled visit estimates assumed eight minutes of exposure, not a universal appointment duration. Under that assumption, the median dose was about 5.1 J/cm² per visit. Depending on the lamp, they calculated 8–208 visits to accumulate 60 J/cm², a comparison benchmark drawn from earlier DNA-damage work. Shipp et al., JAMA Dermatology.
Those figures describe a measurement-and-dose calculation, not a clinical threshold separating safe from unsafe manicures. They do not mean the seventh visit is harmless or the eighth causes cancer. The calculation does not establish how repeated exposures, biological repair and a particular person’s other UV exposure combine to affect risk. It also sampled older devices, so its results cannot certify the output of a current home or salon lamp.
Diffey’s 2012 mathematical model estimated that tens or hundreds of thousands of regular users would be needed for one additional squamous cell carcinoma on the backs of the hands. This was an estimate built from age and exposure assumptions, not an observed rate in a cohort of manicure clients. It supports a low-risk interpretation under the model’s conditions, while leaving the assumptions and their applicability open to scrutiny. Diffey’s risk model.
A 2018 review by Shihab and Lim likewise characterized the risk as low on the evidence then available, while recommending broad-spectrum sunscreen. Its publication predates the newer cell experiments. Older reassuring assessments belong in the discussion, but they should not be presented as measurements of contemporary clients’ lifetime risk. Shihab and Lim’s review.
Case reports and registry trends leave different gaps
MacFarlane and Alonso’s 2009 report described two middle-aged women who developed nonmelanoma skin cancers on the backs of their hands after UV nail-light exposure. Such observations raise a question worth investigating. With two cases and no comparison group, however, they cannot isolate the effect of nail lamps from other lifetime UV exposure or establish how common the outcome is. Original case report.
A case report identifies an event and a possible exposure history. It supplies neither the denominator of all exposed users nor a comparable unexposed group. Without those, it cannot calculate an excess risk. This limitation is a reason to avoid claiming causation from the report; it is also not a reason to ignore the cases. They remain part of the evidence that motivates better research.
Schwartz and colleagues’ 2020 paper found no matching published cases in its search for patients under 40 with chronic gel-manicure histories and hand or nail-unit cancers. The search was conducted in April 2019. The paper also reported no change in melanoma incidence among people under 65 during 2007–2016. These figures describe the study’s historical observation windows; they are not a current surveillance estimate for lamp users. Schwartz et al., 2020.
Neither result tests cancer risk by recorded nail-lamp exposure. A literature search can miss unreported or unrecognized cases, and broad melanoma trends do not answer the same question as hand squamous cell carcinoma after manicures. Interpreting these data as proof that registries would already have detected any hazard goes beyond the study design. The appropriate conclusion is that these analyses did not establish the exposure-specific answer.
An evidence ladder for reading the claims
This table compares what each design measures. It is not a ranking in which the later rows cancel the earlier ones. A review can organize several kinds of evidence, but it cannot create the missing human follow-up data.
| Design | Examples and exposure context | What it measures | What remains unanswered |
|---|---|---|---|
| Cell experiments | Zhivagui 2023: prolonged exposures; Słabicka-Jakubczyk 2023: four versus twenty minutes. | Cell survival, oxidative injury or mutations. | Show a hazard under tested conditions; do not measure a client’s cancer probability. |
| Exposure and heat controls | Finn 2024: salon-style and prolonged schedules; temperature controls. | Different cellular endpoints and heat effects. | No viability loss in one cell type does not mean no injury in all cells or intact skin. |
| Lamp measurements and models | Shipp 2014; Diffey 2012. | Output, estimated dose and modeled excess cancers. | Depend on sampled lamps and assumptions; do not set a safe visit allowance. |
| Human case reports | MacFarlane and Alonso 2009. | Cancer diagnoses after reported exposure. | No comparison group or exposed-population denominator. |
| Literature search and registry context | Schwartz 2020. | Published young-adult cases and broad melanoma trends. | No exposure-specific comparison; absence of a signal cannot establish safety. |
| Evidence syntheses | Metko 2024; Beylin 2025. | Patterns and gaps across heterogeneous studies. | Inherit weaknesses of the included research and the limits of their search dates. |
What the 2024 and 2025 reviews concluded
Metko and colleagues screened 2,331 nonduplicate records and included nine studies: five experimental studies, one cross-sectional study and three case reports. Risk of bias was high or unclear. Their conclusion was that prolonged, repeated exposure may pose a low risk, with weak evidence that patients need to understand when making decisions. They recommended discussing UV-blocking gloves or properly applied sunscreen. 2024 systematic review.
Beylin and colleagues’ 2025 scoping review included 17 articles and searched through January 2024. It reported generally low typical-use risk assessments, variable lamp emissions and some exposures exceeding safety limits. It found no conclusively established causal link. It also judged the evidence insufficient to mandate gloves or sunscreen, a policy conclusion that does not mean optional protection is ineffective or unnecessary for every person. 2025 scoping review.
The January 2024 search cutoff matters: a review’s publication date is not the date through which it captures evidence. Neither synthesis supplies a reliable individualized risk estimate. A reader can accept that uncertainty while choosing a precaution: reducing an optional exposure does not require proving that it caused a particular cancer.
US cosmetic oversight is not a lamp safety certificate
FDA’s Nail Care Products page explains that nail cosmetics generally do not need premarket approval, except for most color additives. Products must comply with safety requirements. State and local authorities oversee salon operations and licensing, and OSHA addresses worker safety. The page concerns cosmetic products and ingredients; it does not provide a curing-lamp emission limit or certify that a particular lamp prevents cancer. FDA Nail Care Products.
The distinction helps when a salon or product listing uses regulatory language. A statement about a polish formula, a salon license or a worker-safety rule answers a different question from the biological effects of a curing lamp. The absence of lamp specifications on that consumer page also cannot establish that no federal requirements of any kind apply to the equipment. Ask what an approval or safety claim actually refers to.
Choosing protection or a service without a lamp
MD Anderson dermatologist Saira George characterizes the apparent risk as low while advising people to reduce avoidable exposure, particularly salon regulars and people with fair skin, substantial past UV exposure or previous skin cancer. The cancer center suggests sunscreen on exposed skin, fingerless gloves, or services using regular or dip polish that do not need UV curing. Confirm the actual service, including its finishing coat, avoids the lamp. MD Anderson’s advice.
Sunscreen: choose the broad-spectrum, water-resistant SPF 30+ protection recommended by the AAD and use it according to its label before exposure. Plan with the manicurist so skin preparation does not inadvertently remove it before curing.
Gloves: dark, opaque fingerless gloves are an AAD option. Look at what they actually cover; openings that expose the fingertips and nail folds leave those areas uncovered. A glove is a barrier for covered skin, not a promise that the entire hand receives no UVA.
Frequency: reserving gel manicures for fewer occasions reduces the number of lamp exposures. No cited study establishes a universally safe weekly, monthly or lifetime schedule. Do not shorten the product’s required cure as a substitute for choosing a service without a lamp.
The laboratory sunscreen experiment placed SPF 50 cream on a plate lid, rather than testing sunscreen on clients’ hands. Improved cell survival in that setup supports UV attenuation but cannot quantify a reduction in human cancer risk. The same paper summarizes evidence that the nail plate transmits little UVA, while the surrounding skin remains exposed; that is not proof of immunity for the nail unit. Sunscreen experiment and nail-transmission discussion.
Avoiding a lamp addresses the radiation question only. It does not make another nail service free of allergy, irritation or nail damage. If symptoms rather than cancer headlines are driving the decision, bring that problem to a clinician rather than assuming a switch of lamp or polish will solve it.
When the evidence cannot answer your personal situation
The studies summarized here do not provide dependable separate cancer-risk estimates for Fitzpatrick IV–VI skin, frequent home-lamp use or a technician’s working day. Do not infer zero risk in darker skin or assign the same numerical risk to every skin tone. Occupational exposure needs its own assessment rather than multiplying a client model by working hours.
If a medicine or a medical condition makes you sensitive to light, ask your prescriber or dermatologist how that changes the decision before an appointment. A generic list of drug classes cannot determine whether your particular treatment is compatible with the lamp. Do not stop prescribed medication to keep a cosmetic appointment.
Seek dermatology assessment for persistent rough or scaly areas, a sore that does not heal, or a changing hand or nail lesion. A previous manicure neither diagnoses the cause nor rules out another problem. The AAD advises assessment for unusual nail changes or persistent nail problems. Arrange evaluation rather than covering a persistent or changing lesion with another manicure. AAD advice on persistent nail problems.
For a rash after a manicure, the separate guide to eyelid reactions and nail acrylates covers a different exposure pathway. Bring the products used and the timing of symptoms to the consultation.
