UVA and skin aging: how photoaging actually happens

Most of what people call ageing skin is not ageing. Compare the skin on the inside of your upper arm with the skin on the back of your hand: same person, same number of years, radically different texture, tone and firmness. The difference is not time. It is accumulated ultraviolet, and the band doing most of the structural damage is UVA.

Why depth decides everything

Skin ages structurally in the dermis: the layer beneath the epidermis where collagen fibres provide tensile strength and elastin provides recoil. Wrinkles, laxity, crepiness and loss of firmness are all dermal phenomena. Whatever reaches that layer is what ages you.

This is where UVA and UVB part company. Absorption in tissue is strongly wavelength-dependent, and longer wavelengths penetrate further:

Now combine that with abundance. At the ground, UVA outnumbers UVB by roughly twenty to one. A band that is both twenty times more plentiful and several times more likely to reach the dermis delivers overwhelmingly more dermal dose — which is why photoaging is fundamentally a UVA story even though sunburn is a UVB one.

The mechanism, photon to wrinkle

UVB damages DNA directly: a photon is absorbed by adjacent pyrimidine bases and fuses them. UVA works differently, and the indirect route it takes is precisely what makes it a structural problem.

1. Absorption by chromophores. UVA is absorbed not by DNA but by other molecules in the skin — riboflavin, porphyrins, melanin precursors, components of the mitochondrial electron transport chain.
2. Reactive oxygen species. Those excited molecules pass their energy to oxygen, generating singlet oxygen, superoxide and hydroxyl radicals. This is oxidative stress, generated in bulk, deep in the tissue.
3. Signalling cascade. Reactive oxygen species activate cell-surface receptors and the MAP kinase pathway, which converges on the transcription factor AP-1.
4. Collagen-degrading enzymes. AP-1 upregulates the matrix metalloproteinases — chiefly MMP-1 (interstitial collagenase), plus MMP-3 and MMP-9. MMP-1 cleaves collagen types I and III: the exact fibres that hold skin taut.
5. Synthesis is suppressed at the same time. AP-1 also interferes with TGF-β signalling, reducing new procollagen production. The skin is dismantling its scaffolding while building less of it.
6. Incomplete repair, repeated. A single exposure causes a transient burst. The damage is largely repaired — but not perfectly. Each cycle leaves a little disorganised, fragmented collagen behind, and decades of daily exposure accumulate into visible change.

Alongside collagen loss, damaged elastic fibres accumulate as solar elastosis — a build-up of abnormal, non-functional elastotic material that gives severely photoaged skin its characteristic thickened, yellowish, leathery quality. Unlike collagen, this material is not meaningfully cleared.

The evidence

Occupational asymmetry

The most-reproduced image in this field is a 2012 New England Journal of Medicine clinical photograph of a 69-year-old American truck driver who had spent 28 years with his left side toward the window. The left half of his face shows deep furrowing, thickening and coarse wrinkling; the right half looks decades younger. It is the same man, the same age, the same genetics — differing only in cumulative exposure through side window glass, which transmits little UVB and a great deal of UVA.

The pattern generalises. Studies of drivers in right-hand-drive and left-hand-drive countries find skin cancers and photoaging concentrated on the driver's side, and the asymmetry reverses with the driving side.

Twins

Studies of identical twins discordant in sun exposure and smoking consistently find that the more sun-exposed twin is judged older, with more wrinkling and pigmentary change, despite an identical genome. It is as close to a controlled experiment on human ageing as ethics allows.

A randomised trial

The strongest evidence is the Nambour skin-cancer prevention trial in subtropical Queensland. 903 adults were randomised to daily broad-spectrum sunscreen or to discretionary use, and followed for four and a half years, with skin ageing assessed objectively by microtopography of silicone casts taken from the back of the hand.

The daily-use group showed no detectable increase in skin ageing over the period, and roughly 24% less ageing than the discretionary group. These were middle-aged adults with substantial existing sun damage — meaning the intervention worked even when started late.

What this changes in practice

• The exposure that ages you is mostly incidental. Not beach holidays — commuting, walking to lunch, sitting near a window, the school run. It is low-intensity, unremarkable and constant, which is exactly why it goes unmanaged.
• Winter and cloud are not exemptions. Because the UV Index collapses in winter while UVA falls only modestly, the ratio of ageing dose to warning signal is worst in the months people stop thinking about sun. See seasonal UVA and UVA on cloudy days.
• SPF alone does not tell you what you need. A high-SPF product with weak UVA filtration prevents burning while permitting the dermal dose that drives photoaging. What matters is the UVA rating — the circled UVA mark, PA+++ or higher, or a broad-spectrum claim. Decoding UVA labels covers how to read them, and applying sunscreen properly covers why the label figure is rarely what you get.
• The car and the desk are real exposures. Side windows are typically untreated glass. A window seat is a daily UVA dose with no burn and no cue.

What actually repairs existing damage

The honest framing is that repair is partial and prevention is disproportionately effective. Collagen can be rebuilt to a degree; elastotic material largely cannot. Every unit of UVA you do not receive is a unit you never have to treat — and knowing the current UVA Index is a reasonable way to know when it is worth the effort.

Photoaging is one of several distinct harms UVA causes; the others — indirect DNA damage, skin-cancer risk, immune suppression — are covered in the dangers of UVA radiation.