What UV radiation does to the skin: why it harms, where it helps, and what science knows about skin cancer

I'm studying to be a beautician and the in an article about a minimum skin care routine I wrote what the minimum routine is, which everyone does well: cleansing, moisturizing, sun protection. The third step, sun protection, is part of the routine because it protects against the sun's ultraviolet (UV) radiation. What this radiation does to the skin, at the cellular level and in the long term, is presented in this article.

In the spectrum of sunlight, two UV bands reach our skin: UVB (280-320 nanometers) and UVA (320-400 nanometers), and a part of visible light (400-760 nanometers) also plays a role. A nanometer is one millionth of a millimeter, that's the wavelength with which the bands are separated. According to Yang and Lan's 2025 review, UV radiation causes direct DNA damage: if the cell repairs the damage incorrectly, a genetic mutation is left behind. In addition, UV also produces reactive oxygen radicals (ROS, aggressive molecules formed from oxygen), which trigger inflammation and immunosuppression (weakening of the immune system), all of which contribute to cancerous transformation.1 This is a summary statement, not a specific measurement, and does not give an exact % of the chance that a mutation will result in a cancerous cell.

Blue light (400-500 nanometers) is not UV, but it is a related topic because it is part of sunlight. McKay and Wu's 2026 review collected lab studies in which blue light caused oxidative stress and slowed the repair of UV damage in a skin model; the authors found no proven direct link with cancer.2 According to the guidelines of the International Commission on Radiation Protection (ICNIRP), visible light does not increase the risk of skin cancer.18 The effect is also measured on humans: blue and violet light can cause pigmentation on skin type III (sometimes burns, but gradually tans) and darker skin; in one study, after 415 nanometer light, the stain remained for up to three months.19

This is also interesting because of the LED masks, because according to the manufacturer's measurement, the blue chip of my mask lights up around 414 and 415 nanometers, in the blue, purple, cyan and white modes, exactly where the above test measured the spot; pigmentation developed in a double-blind study with a 450 nanometer LED.20 A Cochrane review of phototherapy treatments for acne rated the evidence on side effects as very low quality, and less than half of the studies looked at participants more than eight weeks after the end of treatment, so little is known about long-term effects.21 If your skin is gradually browning (III) or darker, or you are prone to spotting (melasma, dark spots after inflammation), I think it is worth using the methods using blue tiles (blue, purple, cyan, white) with caution.

What you can take away from this: UV directly damages skin cell DNA, and this damage is the common root of the two things this article is about: skin aging and skin cancer.

Tanning (increased melanin production) is presumably the skin's defensive response to UV-induced cell stress. This is my own reading, not a literal statement from the Yang and Lan review: the review describes DNA damage and reactive oxygen species as the mechanism of UV-caused skin cancer, but doesn't discuss tanning as a defensive response.1 A more direct, but different kind of source is Cui and colleagues' 2007 experiment, published in the journal Cell: on mouse skin and cell culture, they showed that UV-induced pigmentation is driven by the p53 protein (one of the cell's central regulators responding to DNA damage).10 This is animal and cell-experiment data, not evidence directly about humans, for the tanning-DNA-damage link.

What you can take from this: tanning may signal that cells have responded to a UV load, but neither of these two sources states whether the speed or depth of tanning in a given person is proportional to the extent of DNA damage.

Yes, two, and both come with a limit worth mentioning alongside them. The first is vitamin D: sunlight's UVB radiation is what starts vitamin D3 synthesis in the skin. According to Uçar and Holick's 2025 review, the same UVB radiation that triggers vitamin D production can, in excess, also contribute to skin cancer (both non-melanoma and melanoma) through DNA damage and oxidative stress; this duality is the underlying problem in balancing sun protection against vitamin D status. The authors state that more than half of the world's population is at risk of vitamin D deficiency or insufficiency, although this review doesn't detail the primary epidemiological source or the threshold behind that specific figure.3

The question of whether sunscreen worsens vitamin D status is answered by an international expert panel's position statement, with this important background: Passeron and 14 co-authors published a position statement in the British Journal of Dermatology in 2019, based on a 13-member expert panel's (endocrinologists, dermatologists, photobiologists, epidemiologists and anthropologists) literature review and one-day consensus meeting; the review was funded by L'Oréal, and five of the 15 authors are employees of the company. The position statement holds that broad-spectrum sunscreens that prevent erythema (skin reddening) are unlikely to worsen vitamin D status in a healthy population, and that everyday, recreational sunscreen use doesn't impair vitamin D synthesis, even under optimal application conditions. The "healthy populations" qualifier matters: a different recommendation applies for photosensitive patients, where vitamin D screening and supplementation is advised.4

The second good side is one you can feel in everyday life: daylight sets your internal clock. There's a special light-sensitive cell type in the eye's retina that signals to the body when it's daytime, independent of vision.16 In a small trial, eight people lived in natural light only for a week in a campsite, and their internal clock shifted to align with sunrise and sunset, roughly two hours earlier.5

It connects to mood too. In a large UK dataset, people who spent more time outdoors during the day reported feeling low or having insomnia less often; this is an association, it doesn't tell you which one causes the other.14 For seasonal, winter-linked low mood, bright light therapy helped moderately according to studies, though these trials are small and of mixed quality.15

And here's the best part: the known pathway for this effect is visible light reaching your eyes, not UV reaching your skin. According to an international clinical recommendation, light therapy's effect doesn't require UV, which is why only UV-filtered lamps are recommended, and a 2026 public health summary states outright that you can seek out daylight alongside sunscreen too.1617 So you don't need to get sunburnt for a better mood.

What you can take from this: sunlight has two real benefits, vitamin D, and your internal clock and mood. For vitamin D your skin needs UV, for the other, it's enough for your eyes to get daylight, which is why experts say you can get the latter alongside sunscreen too; I didn't find a direct measurement for this.

The best data on this comes from an Australian trial, and the answer depends on the type, not a simple yes. 1621 residents of the Queensland town of Nambour, aged 25 to 75, were randomised into two groups: one applied sunscreen daily for 4.5 years, the other according to their own habit, occasionally. Green and colleagues published the trial in the Lancet in 1999. Beta-carotene supplementation was also tested alongside it, and only cancers appearing on the daily-treated body sites (head, neck, arms, hands) were counted. For squamous cell carcinoma (SCC, a skin cancer arising from the upper layers of squamous cells), fewer new tumours developed in the daily sunscreen users: 1115 per 100 thousand people, versus 1832 in the occasional group. This difference was statistically reliable, meaning it's unlikely to be down to chance alone. This, though, applies to the number of tumours. For how many people developed a new SCC at all, there was no detectable difference between the two groups. The authors' own wording is cautious too: SCC "appears to be preventable" with sunscreen, they don't state this as proven.6

For basal cell carcinoma (BCC, a skin cancer arising from the lowest, basal cell layer of the epidermis) there was no effect: neither in the number of tumours, nor in how many people developed a new BCC at all (this latter was 2588 per 100 thousand people, versus 2509, practically the same). This is an important counterbalance: sunscreen's cancer-protective effect is type-dependent, and this trial showed no effect for BCC.6

For melanoma, the long-term follow-up gives the answer. Green and colleagues, in a follow-up published in the Journal of Clinical Oncology in 2011, looked at the same participants 10 years after the trial ended, so roughly 14 years after it started (from 1992 to 2006). The daily sunscreen group had 11 new primary melanomas identified, versus 22 in the occasional group. This difference sits right at the edge of statistical reliability: the calculation's uncertainty range just crosses over the point where there would be no difference at all, so chance can't be ruled out. For invasive melanomas (already spread into the skin's deeper layer), the reduction was statistically detectable, but from very few cases: 3 cases in the daily group, 11 in the occasional group, and the edge of the uncertainty range here, although it doesn't quite reach the "no difference" value, comes close to it. The authors' own wording is cautious here too: they state that melanoma may be preventable with sunscreen, without claiming this as proven themselves.7

The Cochrane (an independent, international network producing systematic reviews) 2016 review adds context here: keratinocyte-derived skin cancers, meaning those arising from the skin's main cell type, (BCC and SCC combined) make up around 95 percent of malignant skin tumours, and new cases worldwide increase by somewhere between 3 and 8 percent a year. This proportion is about the split between skin cancer TYPES, showing how large a share is keratinocyte-derived compared with melanoma. This is not the same claim as saying sunbathing causes 95 percent of cancer cases, and it's important not to conflate the two. The Cochrane review adds an important limitation here too: they found only this one trial (the Nambour trial) suitable on the topic, found no difference in per-person incidence for either BCC or SCC, and rated this evidence itself as low certainty, partly because some of the cancerous lesions weren't confirmed histologically.8

What you can take from this: daily sunscreen use produced fewer SCC tumours in this trial, but had no detectable effect on whether someone developed SCC at all. For basal cell carcinoma, this trial showed no difference whatsoever. For melanoma, the long-term reduction was detectable for invasive cases, but based on few cases, while for the whole group it sits at a borderline result. An independent review rated this same evidence as low certainty.

The WHO's International Agency for Research on Cancer (IARC), based on an expert working group's evaluation in July 2009, classified UV-emitting tanning devices in Group 1 (carcinogenic to humans), and the decision was published in the journal The Lancet Oncology.9 The IARC's four categories (1, 2A, 2B and 3) express the strength of carcinogenic evidence, with 1 being the strongest.13

A lot of people ask whether there's such a thing as a healthy sunbed, and how much sunbed use is okay. I didn't find a study measuring any beneficial effect from sunbeds, nor one giving a safe dose. My own view is there's probably no such thing as a healthy sunbed. In the cold, warmth just feels nice, of course, and if someone goes for the mood, the section above suggests daylight reaching your eyes is the better route, without UV; you can get warmth without UV radiation too.

What you can take from this: the WHO's cancer research agency classifies tanning devices as a proven human carcinogen, and I found no scientific support for the claim that there's a beneficial or safe sunbed.

Darker skin provides natural protection against UV, but this protection isn't complete. It has more dark-brown pigment (eumelanin), and the pigment-storing cell granules (melanosomes) are more mature. According to Leybengrub and colleagues' 2026 literature review, skin cancer INCIDENCE is lower in dark skin, but MORTALITY is higher, likely due to late or misdiagnosis; people with darker skin types can also get UV damage, pigmentation disorders, photo-induced skin lesions and skin cancer.11 A related methodological problem: according to Olsen and colleagues' 2026 review, the Fitzpatrick scale (a self-reported, grade I-VI system classifying skin's response to sunburn/tanning) has received considerable criticism, because it relies on self-reporting, the questions asked vary, its range is limited, and it UNDERESTIMATES skin cancer risk in certain groups.12 So the claim that "people with dark skin don't need sunscreen" is false: the risk is lower, not zero, and the self-reported estimate of skin type can also be inaccurate.

The cloud and window glass question can be answered at the level of a physical principle, without a specific figure: cloud only filters out part of UV radiation, a significant share still gets through. I didn't find a precise percentage for this, so I'm not giving a number; common sense says that if cloud doesn't absorb all of it, it's not worth skipping sunscreen on a cloudy day either. For window glass, the physical principle is that normal (non-UV-filtered) glass absorbs most UVB radiation, but lets a significant share of UVA through, so indoors, next to a window (in a car, in an office), exposure is mainly to UVA. I didn't find a precise measurement for this either (for example, how much UVA passes through a car windscreen), so this stays a physical principle, without a number. Common sense says that if you spend a lot of time next to a window or in a car, this UVA load adds up day after day.

What you can take from this: darker skin reduces but doesn't eliminate the risk, and cloud and window glass only partially filter UV, so sunscreen is warranted in both situations, even though the exact percentage can't be backed by a source today.