How to stop skin aging: what the studies back and what they don't

Sugar really does stiffen the collagen in your skin, but no study has shown that eating less of it makes your face look younger. The chemistry behind the claim is solid, while the payoff on your plate is not. The far bigger driver of visible facial aging is the sun.^[11]

That chemistry is called glycation. Sugar sticks to proteins without any enzyme steering the reaction and, over time, turns irreversibly into compounds called AGEs (advanced glycation end-products).^{[1, 2]} In skin, AGEs link the strands of collagen and elastin together, the fibers that keep skin firm and springy. Those cross-linked fibers stiffen, lose elasticity and add to the dull, yellow-brown tone of older skin.^{[1, 2]}

Stiffer fibers are only half of the damage, because AGEs also disturb how collagen is renewed. They push up the enzymes that cut collagen apart and slow the making of new collagen.^{[1, 2]} Heavily cross-linked collagen is also harder for those enzymes to clear, at least in lab and mouse studies.^[30] So old, damaged collagen stays in the skin, and less fresh collagen arrives to replace it.

That slow clearance is why AGEs pile up over decades. Skin collagen is extremely long-lived: its half-life, the time it takes for half of it to be broken down, is roughly 15 years.^[3] Its cross-links only disappear when the protein itself is cleared.^[3] That store can even be measured without a needle. A device called an AGE Reader measures how strongly the skin glows under light, and that glow rises with age, smoking and waist size.^{[2, 20]}

Whether food changes that store is far less clear, and we found no study showing that eating less sugar lowers skin AGEs. Even for the AGEs you eat directly, the browned compounds in fried and roasted food, the link is weak. In the Rotterdam Study, a large Dutch population study, food AGEs showed no link with the skin's glow across the whole group.^[4] A weak link appeared only after the researchers left out people with diabetes and kidney disease, and it concerned just one type of AGE.^[4] The authors read that as a hint that food AGEs can add to the skin store.^[4]

A steadier blood sugar is still a sensible habit for general health, and the guide on glucose and blood sugar hacks covers how to keep it flatter. But no study has shown that it slows skin aging in healthy people, and the Rotterdam result does not support avoiding browned food either.^[4] The only step around glycation with trial evidence for skin is daily UV protection, which shields collagen from the outside.^[14]

Collagen peptides are small pieces of collagen sold as powders and drinks, and they show benefits mainly in trials paid for by their makers.^[23] Independent, high-quality trials find no effect on the skin.^[23] So their benefit is unproven, not just small. Creams you put on your skin are a different story, because prescription retinoids have far more and longer trials behind them.^[9] Vitamin C sits in between, with a sound rationale and a handful of small trials.^[10]

Collagen powder: good early results, weak trials

The early collagen trials did look promising. Compared with a dummy powder, 1 to 5 g of collagen a day made skin measurably more elastic after 8 to 12 weeks.^{[5, 8]} One of these trials also found about a fifth less wrinkle volume around the eyes.^[6] A meta-analysis, which pools the results of many trials, then found more hydration and elasticity and fewer wrinkles after about 90 days.^[7] Two of the early trials, though, were funded by the maker of the peptides they tested.^{[5, 6]} Most of their results were also skin measurements rather than how old a face looks.^{[5, 6]}

Who paid for a trial turned out to predict its result. A newer meta-analysis of 23 randomized trials with nearly 1,500 participants sorted them by funding and by quality.^[23] Only the company-funded and lower-quality trials showed benefits, while the independent and high-quality ones found no effect on hydration, elasticity or wrinkles.^[23] Its authors concluded that there is no clinical evidence to support collagen supplements against skin aging.^[23] A later reanalysis of the earlier meta-analyses points the same way. It found misreported data and badly pooled measures, and after accounting for dose, no evidence of benefit.^[24]

Collagen powder is not cheap either, at typical German prices of about 20 to 40 euros a month (as of October 2026). That is a fair sum for an effect the independent trials have not found. On safety, one 12-week trial reported no side effects.^[8] Fish collagen, though, has caused anaphylaxis, a severe allergic reaction, in people whose immune system was already primed against it.^[29] So check the source animal on the label, and if you have a fish allergy, ask a doctor first.^[29]

Retinoids: the best-studied cream against wrinkles

Retinoid creams are where the stronger wrinkle evidence sits. A review of 7 randomized trials looked at tretinoin, a prescription form of vitamin A acid.^[9] It consistently improved wrinkles, blotchy pigment, sallowness (a dull, yellowish tone) and age spots.^[9] Only two of those trials compared it with an inactive placebo cream, while the others tested it against other creams or doses.^[9] The first changes, in coarse wrinkles, showed after about a month in one trial, but most effects took several months.^[9] In the longest trial, the benefit held for 24 months of continued daily use.^[9]

Tretinoin has close relatives, and in the trials that compared them head to head, adapalene and tazarotene worked at least as well.^[9] In one of those trials, tazarotene did somewhat better, mainly because it worked faster.^[9] All of them switch on collagen production and block the collagen-cutting enzymes, so they work against the same damage that sugar and sun cause.^{[9, 12]} Of these creams, tretinoin is still the best-studied against existing wrinkles.^[9]

Tretinoin also comes with rules. It is prescription-only in Germany,^[26] irritates at first (flaking, redness) and needs daily sunscreen alongside it. For wrinkles you usually pay for it yourself, because statutory health insurance excludes drugs used mainly to improve quality of life.^[27] (as of October 2026). Pregnancy is the hard limit. The EU medicines agency rates all prescription retinoid creams and gels (tretinoin, adapalene, tazarotene, isotretinoin) as off-limits in pregnancy and for women planning one^[22] If you are pregnant, planning a pregnancy or breastfeeding, ask your doctor before using any retinoid product, including cosmetic retinol.

Cosmetic retinol is the over-the-counter cousin. The skin converts it into the active acid, but it has far fewer trials behind it. In one trial in the review, a retinol combination cream did about as well as a low-strength tretinoin cream and was better tolerated.^[9] That is one product, though, not proof that retinol matches tretinoin.

Retinol in cosmetics is now also capped, because vitamin A from food, supplements and cosmetics adds up.^[25] Since November 2025, new cosmetics in the EU may contain at most 0.3% retinol equivalent (0.05% in body lotion), and they must carry the label "Contains Vitamin A. Consider your daily intake before use".^[25] From May 2027 that applies to everything on the shelf (as of October 2026).^[25]

Vitamin C: a sound idea with small trials

Vitamin C applied to the skin has a good rationale but only small, early trials. Your skin needs it as a helper for the enzymes that stabilize fresh collagen fibers, and without it stable collagen cannot form.^[10] The trials that test it are tiny, though. In one, a vitamin C patch covered in microscopic needles reduced crow's-feet wrinkles over 12 weeks.^[10] Another trial, with a 5% cream, gave a more mixed picture.^[10] The genes that build collagen became more active, but the skin did not clearly gain more collagen.^[10] The products are also hard to keep stable,^[10] so vitamin C is a reasonable extra, not a replacement for a retinoid.

Read side by side, the four wrinkle levers in this guide split cleanly: two have solid trial evidence, two do not.

Prices are typical German prices and the insurance status is as of October 2026; neither comes from the studies. Which product fits your skin is a conversation with your dermatologist or GP.

Most visible facial aging in fair skin seems to come from the sun, about 80% by the best-known estimate.^[11] That share is why daily sunscreen ranks next to tretinoin as a wrinkle lever. For most of the signs people notice in the mirror, UV light seems to matter more than calendar age.

The 80% comes from a study that graded 298 fair-skinned women aged 30 to 78 for visible aging signs.^[11] Pigment spots were the most sun-driven sign, wrinkles came from both sun and time, and sagging tracked age alone.^[11] The study compared sun-seeking with sun-avoiding women at one point in time, and it was run by cosmetics-industry researchers.^[11] So take 80% as the authors' rough estimate, not a measured share you can prevent. Still, future sun exposure is the biggest factor here that you can change.

The sun's share of the damage has a name: photoaging. UV light switches on collagen-cutting enzymes in the skin, and a classic experiment showed collagen breakdown rising after exposure.^[12] Inner (intrinsic) aging runs everywhere in the body and cannot be halted.^[13] Photoaging, by contrast, happens only where light reaches, mainly the face, neck and backs of the hands.^{[12, 13]} You can see the difference on yourself. The inside of your upper arm has spent most of its life under a sleeve, and it often looks younger than your face, although both are the same age.

The light that does this comes in two kinds, and protection needs to cover both. UVA makes up about 95% of the UV that reaches the ground and penetrates into the dermis, the deeper skin layer that holds your collagen. UVB, the other 5% or so, stays in the outer layer, where it damages DNA and causes sunburn. Both contribute to photoaging,^{[9, 13, 19]} which is why you need broad-spectrum protection, not just a high SPF.

Broad-spectrum sunscreen, used daily, does slow skin aging, as a randomized trial in 903 Australian adults under 55 showed.^[14] Over 4.5 years, skin aging progressed less in the daily-sunscreen group than in people who used sunscreen when they felt like it.^[14] The authors put that difference at 24%.^[14] The daily group, in fact, showed no detectable increase in skin aging over the trial.^[14] The effect is modest, though, some data were missing, and the trial was not large enough to pin down moderate effects precisely.^[14]

That 24% describes slower new damage, not repair. The trial shows that sunscreen slows further photoaging, but it did not establish whether it also reverses existing wrinkles.^[14] It was also run in subtropical Australia, in people under 55.^[14] For wrinkles already there, the trial evidence sits with retinoids.^[9]

In daily life, the American Academy of Dermatology recommends a broad-spectrum, water-resistant sunscreen with at least SPF 30, which blocks about 97% of UVB rays.^[15] Use enough: at least a teaspoon for the face and about a shot glass for the body. When outdoors, reapply about every two hours and after swimming or sweating.^[15] A daily facial SPF costs around 10 to 25 euros over the counter (typical German prices, as of October 2026). By our reading, that makes it the best value of all the levers in this guide.

Skin ages fast around menopause because estrogen, one of the main supports of skin collagen, drops away.^[17] Older studies report about 2% less skin collagen for every year after menopause.^[17] Separate estimates put the loss at up to about 30% in the first five years after menopause.^[17] Both figures come from small studies in the 1980s and 1990s, so they are ballparks rather than exact values.^[17] Still, they fit the abrupt change many women notice.^[17]

That the loss follows hormones rather than birthdays was shown early. In women without hormone treatment, skin collagen fell with the years since menopause rather than with calendar age.^[21] A follow-up then found that estrogen therapy raised it again.^[16] The link sits in the skin itself. Both the cells of the outer layer and the cells that build collagen carry receptors, docking sites that let estrogen act on them.^[17] Through those receptors, the hormone boosts collagen and elastic fibers and slows collagen breakdown.^[17] It also raises hyaluronic acid, a molecule that holds water in the skin.^[17] When estrogen falls, skin thickness, moisture and elasticity fall with it.^[17]

Those receptors are also why hormone therapy can change the skin. Systemic menopausal hormone therapy (MHT), the kind that acts on the whole body, is prescription-only.^[26] and can raise skin collagen and thickness^[17] But guidelines name hot flushes, vaginal and urinary symptoms and bone protection as the reasons to use it, not skin.^[28] Its risks depend on the type, dose, route and timing.^[28] When it is started after 60 or more than 10 years after menopause, the risks of heart disease, stroke, blood clots and dementia weigh more.^[28] So skin should never be the reason to start it. Decide with your doctor, based on your symptoms and your overall risk; the guide on menopause symptoms covers the treatment options in full.

Estrogen creams for the face are a grey zone. Their off-label use, outside what they are approved for, has grown fast.^[18] Yet a recent review found the data thin and the safety questions unsettled.^[18] Small studies show more collagen and a thicker outer skin layer, but safety and benefit are far less well documented than for vaginal estrogen.^{[17, 18]} In Germany estradiol and estriol creams are prescription-only,^[26] so this belongs with a doctor rather than on a skincare shelf.

You can estimate how old your skin looks with the free Face-Age Test and the age test. Use the result as a baseline rather than a verdict. What you get is an estimate of how your skin appears, not your biological age, and it is a self-check, not a diagnosis.

That appearance is also what every lever in this guide aims at, yet none of their trials measured a face-age estimate. The sunscreen trial graded skin aging,^[14] and the tretinoin trials graded wrinkles and pigment.^[9] So a test like this cannot prove that a product works for you. What it can show is whether the overall impression of your skin shifts while you stick with a routine.

The routine with the best evidence is daily broad-spectrum sunscreen.^{[14, 15]} plus a retinoid after talking to a doctor^[9] Leave out the retinoid if you are pregnant or planning a pregnancy.^[22] Take the test before you start, repeat it after a few months, and read any change in the estimate as a rough signal rather than proof.