Skin & Aging

By Maya Sullivan · Published March 4, 2026 · Updated April 12, 2026 · ~10 min read

The framework — what actually drives skin aging

Visible skin aging is not one process. It is the additive output of four mechanisms, and almost every credible intervention maps onto one of them. Rank-ordered by the size of the contribution and the strength of the evidence:

1. UV exposure (photoaging). Ultraviolet A and B radiation drive dermal matrix metalloproteinase (MMP — enzymes that cleave collagen) upregulation, DNA damage, and elastin disorganization. The Australian sunscreen RCT (Hughes, Ann Intern Med 2013) is the cleanest evidence we have that a topical intervention measurably slows visible aging in humans [Hughes 2013].
2. Glycation. Circulating glucose and fructose form cross-links on long-lived dermal proteins — the Advanced Glycation End-products (AGEs) that stiffen collagen and yellow the dermis (Vierkötter, J Invest Dermatol 2010) [Vierkötter 2010].
3. Inflammaging. Chronic low-grade inflammation accelerates MMP activity, impairs fibroblast collagen synthesis, and is the mechanism behind smoking, sleep deprivation, and metabolic dysfunction all showing up on the face.
4. Intrinsic collagen turnover decline. Type I and III dermal collagen synthesis falls roughly 1% per year after the third decade, accelerating perimenopausally.

Read those mechanisms and the implication is obvious. The four-hundred-dollar serum with a hyphenated peptide on the box is not addressing any of them at a meaningful clinical dose. The five-dollar sunscreen and the prescription tube of tretinoin cost less and have the only strong-tier evidence in the category. Most of what gets sold as "anti-aging" is mechanism-adjacent marketing layered on an ingredient with a study its manufacturer commissioned.

Sunscreen — the foundational anti-aging tool

Sun Protection Factor (SPF) is a UVB-only metric. An SPF 30 product blocks roughly 97% of UVB photons; SPF 50 blocks roughly 98%. The clinical difference between the two is smaller than the difference between applying enough sunscreen and not. Real-world application densities are typically 0.5–1.0 mg/cm² — half the 2 mg/cm² used in regulatory testing — which means the SPF on the bottle is roughly twice the SPF on your face.

The Nambour, Australia trial (Hughes, Ann Intern Med 2013) randomized adults to daily broad-spectrum SPF 15 versus discretionary use over 4.5 years. The daily-use group showed 24% less measurable photoaging by microtopography [Hughes 2013]. This is the only large randomized human trial showing a topical product measurably slows visible aging, and the product was sunscreen.

Two practical points the marketing rarely lands on. First, UVA — the longer-wavelength band responsible for the bulk of dermal photoaging — is poorly captured by SPF. The European labeling system uses UVA-PF (UVA Protection Factor) and requires UVA-PF to be at least one-third of SPF, marked with the circled "UVA" logo. North American labels generally say only "broad-spectrum," which is a weaker promise. Buying a European or Asian filter is the most evidence-based skincare upgrade most people can make.

Second, mineral versus chemical is mostly a values discussion, not a safety one. Zinc oxide and titanium dioxide are inert and broad-spectrum but cosmetically heavier. Modern organic filters (bemotrizinol, bisoctrizole, MCE) are photostable, broad-spectrum, and elegant — and available everywhere except the United States, where the FDA has not approved a new filter since 1999. This is a regulatory failure, not a safety signal.

If you do one thing for skin aging, it is daily broad-spectrum sunscreen with a real UVA-PF. Everything else is a rounding error until that habit exists.

Retinoids — tretinoin tier, retinaldehyde, retinol, and the irritation ladder

Retinoids are vitamin A derivatives that bind nuclear retinoic acid receptors and modulate gene expression in keratinocytes and fibroblasts. The downstream effects — accelerated epidermal turnover, increased dermal collagen synthesis, reduced MMP activity — are the most well-characterized in topical dermatology (Mukherjee, Clin Interv Aging 2006) [Mukherjee 2006].

The category is a tier ladder, not a competition. Potency, side-effect burden, and access tier up together:

• Tretinoin (all-trans retinoic acid) — prescription. Binds the receptor directly. The gold standard. Decades of clinical RCTs.
• Tazarotene, adapalene — prescription (adapalene is over-the-counter (OTC) in some markets at 0.1%). Different receptor selectivity, slightly different irritation profile, similar efficacy ceiling.
• Retinaldehyde — OTC. One enzymatic step from retinoic acid. Lower irritation than tretinoin, real efficacy in head-to-head trials.
• Retinol — OTC. Two enzymatic steps from retinoic acid. Roughly ten-fold less potent per molecule than tretinoin, but well-tolerated.
• Retinyl esters (palmitate, propionate) — three enzymatic steps away. The weakest member of the family and the one most prone to marketing inflation.

The "stronger is better" framing is a trap. The clinical question is which tier you can use consistently without barrier disruption. A patient using tretinoin twice a week because of irritation will see less benefit than a patient using retinaldehyde nightly. Tier up only when the current tier is tolerated.

Collagen and the inside-out claims

Oral collagen peptide trials are a category where the marketing has outrun the data, but the data is also not zero. The Proksch trial (Skin Pharmacol Physiol 2014) randomized 69 women to 2.5 g or 5 g of specific collagen peptides daily versus placebo for eight weeks. The treatment groups showed measurable improvements in skin elasticity by cutometry [Proksch 2014]. The signal is real, the effect size is small, and the trial was manufacturer-supported — which we name, not bury.

The mechanistic story — that ingested di- and tripeptides like prolyl-hydroxyproline survive digestion, enter circulation, and signal fibroblasts — has pharmacokinetic support but the leap to a visible cosmetic outcome is still a small leap. If you are already hitting 1.6–2.2 g/kg of dietary protein, the marginal benefit of an additional 10 g of hydrolyzed collagen is probably small. If you are protein-undernourished, fix the protein first.

Vitamin C is the cofactor that matters here. Prolyl and lysyl hydroxylase — the enzymes that hydroxylate collagen precursor residues so the triple helix can form — require ascorbate. Frank deficiency is the easy case (scurvy). The harder case is whether supplemental Vitamin C, oral or topical, accelerates collagen synthesis at normal serum levels. Topical L-ascorbic acid 10–20% has reasonable photoprotection and pigmentation data; oral mega-dosing for skin specifically does not.

AGEs (Advanced Glycation End-products) connect glycemic control to visible aging (Vierkötter, J Invest Dermatol 2010) [Vierkötter 2010]. The Vierkötter cohort linked fasting glucose to perceived age in a non-diabetic population. The honest framing: most of the dietary-AGE story is correlational, the mechanism is biologically plausible, and the protocol implication ("don't sit at chronic hyperglycemia") happens to be good general advice for reasons that have nothing to do with skin.

The peptide angle — GHK-Cu, Argireline, and the topical evidence base

Topical peptides are the category where marketing inflation is at its worst, and where the editorial work — separating mechanism from manufacturer-funded efficacy claims — is most needed.

GHK-Cu (Copper-tripeptide, glycyl-L-histidyl-L-lysine bound to copper) is the most-studied of the cosmetic peptides. Pickart's body of work (Biomed Res Int 2015 review) lays out an in-vitro and animal mechanism for fibroblast stimulation, wound healing, and antioxidant activity [Pickart 2015]. The wound-healing data is strong. The dermal-cosmetic data in humans is thinner — small, mostly unblinded, often industry-funded. A defensible editorial position: GHK-Cu probably does something, the something is probably modest at cosmetic-formulation concentrations, and the topicals available to consumers are wildly variable in actual delivered concentration.

Argireline (acetyl hexapeptide-8) is marketed as a topical "Botox alternative" because it interferes with SNARE complex assembly at the neuromuscular junction in vitro. Whether enough intact hexapeptide crosses the stratum corneum to reach a motor end-plate at a pharmacological dose is the unsolved question. The published human trials are small, short, and predominantly manufacturer-supported. A reasonable read: the effect, if real, is much smaller than the marketing implies and not comparable to an actual neurotoxin.

Matrixyl, palmitoyl tripeptides, and the next-generation cosmetic peptides share the same pattern. Compelling in vitro signaling, plausible mechanism, small manufacturer-funded human trials, no independent replications at the scale you would want before paying $200 for a 30 mL bottle.

The skincare community has run ahead of the evidence here for years. The editorial framing we use on Wellness Radar: topical cosmetic peptides are not in the same evidence tier as sunscreen or retinoids, and any product priced as though they were is selling brand, not biology.

Lasers, microneedling, biostimulators — when in-office is worth it

For structural changes that topicals cannot deliver — meaningful collagen remodeling, deep wrinkle reduction, volume restoration — in-office procedures are where the evidence tier jumps. They are also where the cost, downtime, and operator-skill dependency jump.

• Fractional ablative CO2 laser. The most evidence-supported single-treatment intervention for moderate-to-severe photoaging. Controlled thermal injury triggers a wound-healing cascade and durable neocollagenesis. Real downtime (7–14 days), real risk of post-inflammatory pigmentation in darker skin types.
• Non-ablative fractional lasers (Fraxel Dual, Clear+Brilliant). Less aggressive, less downtime, less per-session benefit. Stack 3–5 sessions for a result roughly comparable to one ablative pass.
• Microneedling and microneedling-RF (Morpheus8, Vivace). Mechanical and thermal collagen induction. Evidence for textural and scar improvement is solid; the marketing of microneedling as equivalent to laser is generous.
• Poly-L-lactic acid (Sculptra). A biostimulator injectable — not a filler in the volumetric sense. Stimulates gradual collagen deposition over months. Goldie and colleagues have published the longer-term efficacy and safety data [Goldie 2018].
• Calcium hydroxylapatite (Radiesse). Dual-mechanism — immediate volume plus biostimulation. Hyperdilute Radiesse protocols are increasingly used for skin quality rather than volume.

The honest framing on procedures: they work, they cost real money, and they require an operator whose portfolio you can see. They do not replace sunscreen and a retinoid — they layer on top of an established topical baseline.

The boring foundations — sleep, glycemic control, smoking, omega-3, protein

Most of the variance in how visibly people age is not on the bathroom shelf. It is the five inputs that no serum line wants to talk about:

1. Sleep. Chronic short sleep elevates cortisol, suppresses overnight growth hormone pulsatility, and is independently associated with measurable decreases in skin barrier recovery and increases in periorbital hyperpigmentation. The intervention is not a $90 night cream. It is seven to nine hours.
2. Glycemic control. Chronic hyperglycemia drives AGE accumulation in long-lived dermal proteins (Vierkötter 2010). Practical implication: the same glycemic discipline that protects the cardiovascular system protects the dermis.
3. Smoking cessation. Smoking is the second-largest modifiable driver of facial aging after sun exposure. MMP upregulation, microvascular impairment, and direct fibroblast toxicity — all dose-dependent, all partially reversible. The cessation effect on skin is large and visible.
4. Omega-3 fatty acids. EPA and DHA support barrier integrity and dampen inflammatory signaling. Dietary fish twice weekly or 1–2 g daily of a quality-tested supplement is the reasonable target.
5. Protein adequacy. Collagen, elastin, and the enzymatic machinery that repairs them are all built from dietary amino acids. The 1.6–2.2 g/kg reference range that supports muscle synthesis is also the range that supports dermal turnover. Under-protein dieting visibly ages people.

Sunscreen, a tolerated retinoid, sleep, protein, and not smoking. Five inputs explain most of the variance — and four of them are free.