"Anti-ageing" creams won't fix menopausal skin. Here's what actually does.

A cosmetic chemist/biochemist's deep dive into the molecular biology of menopausal skin, the topical oestrogen evidence, and the protocol that actually works.

“Menopause skincare.” It’s now a category.

Premium-priced. Separately positioned. Marketed specifically to women in perimenopause and beyond.

The science of menopausal skin is real. The biology is specific and well-characterised. But most products in this new category don’t reflect that biology any better than well-formulated standard skincare does, and many cost two to three times more for the same actives in different packaging.

Let me say this plainly: your skin in your forties and fifties is doing something specific, predictable, and substantially modifiable.

Oestrogen, which has been quietly running large parts of your skin’s collagen production, lipid synthesis, hydration, and barrier function for decades, is leaving the building.

That’s not a failure. That’s a biology change. And it has a protocol-shaped solution, not a serum-shaped one.

Here’s my promise for the next 3,000 words:

By the end of this letter, you’ll understand what’s actually happening to your skin in perimenopause and menopause at a molecular level. You’ll have a protocol you can build a routine around. And you’ll know which ingredients have the peer-reviewed evidence to genuinely support menopausal skin, and which are marketing.

No “miracle” creams required.

Let’s get into it.

---

The protocol, before the science

Five things are happening simultaneously in menopausal skin:

• Collagen and elastin loss, with the type III/I ratio shifting and dermal thickness decreasing

• Stratum corneum lipid depletion, particularly cholesterol, making the barrier slower to repair

• Reduced ceramide synthesis, with the remaining ceramides shorter and less protective

• Increased oxidative stress and reduced antioxidant capacity

• Altered melanocyte behaviour, worsening melasma, more solar lentigines

No single product addresses all of that.

You need a team. Here’s mine.

---

Step 1: A cholesterol-dominant moisturiser

I’m putting this first because it’s the swap with the strongest mechanistic evidence and the least mainstream attention.

You’ve probably been told to use a “ceramide moisturiser” for ageing skin. That’s not wrong. But it’s not specific enough.

What does ageing actually do to those lipids?

Let me walk you through the human evidence specifically, because this is one of those areas where mouse data and human data don’t fully overlap, and the difference matters.

In aged human skin, total stratum corneum lipid mass declines. Rogers, Harding, Mayo, Banks, and Rawlings (1996) measured stratum corneum lipids in 49 women aged 21–60 using tape stripping at three body sites. All major lipid classes: ceramides, cholesterol, and fatty acids, declined in absolute mass with age, with the decline in facial cholesterol reaching statistical significance. Overall lipid content dropped by roughly 30%.

But here’s the nuance: the relative proportions of the three lipids stayed broadly constant. This isn’t a story of cholesterol being selectively depleted while ceramides hold steady. It’s a story of global lipid loss.

Ghadially, Brown, Sequeira-Martin, Feingold, and Elias (1995) confirmed the functional consequence: in subjects over 80, barrier recovery after disruption was noticeably delayed compared to young subjects, only 15% recovery at 24 hours versus 50% in young skin.

So why “cholesterol-dominant” rather than “any lipid replenishment”? This is where the evidence gets specific.

Ghadially, Brown, Hanley, Reed, Feingold, and Elias (1996), and this one I need to flag carefully, established in aged mice that the rate-limiting deficit isn’t a general lipid problem. It’s specifically cholesterol synthesis.

HMG-CoA reductase activity (the rate-limiting enzyme of sterol biosynthesis) was reduced in aged murine epidermis, while ceramide and fatty acid synthesis upregulated normally after barrier perturbation. This is mouse data. It’s mechanistically clean and biologically plausible for humans, but it has not been replicated as a primary biochemical finding in human aged epidermis with the same rigour.

What has been demonstrated in humans is the functional consequence.

Zettersten, Ghadially, Feingold, Crumrine, and Elias (1997) tested four different topical lipid mixtures on chronologically aged human skin (80 ± 5 years, n=6) after tape-stripping. The cholesterol-dominant mixture (1:3:1 molar ratio of ceramide : cholesterol : fatty acid) significantly accelerated barrier recovery at 6 hours (p<0.005). The fatty-acid-dominant mixture delayed recovery. Equimolar mixtures worked fine in young skin but didn’t add benefit in aged skin, only cholesterol dominance did.

That’s the cleanest human evidence for the recommendation. Small sample, but controlled, peer-reviewed, and species-specific to humans.

One further wrinkle worth knowing about: more recent imaging studies have shown that cholesterol sulfate, a modified cholesterol species, actually accumulates with age in the stratum corneum. So the picture isn’t simply “your skin runs out of cholesterol.” It’s “your skin’s cholesterol metabolism shifts in ways that compromise barrier function, and topically applied cholesterol functionally helps.”

Layer on Kao et al. (2022), which showed the menopausal stratum corneum has lower total ceramide content and shorter ceramide chain lengths, with HRT preventing these changes, and you’ve got a barrier that’s lipid-deficient across the board, with functional evidence pointing specifically toward cholesterol-dominant replacement as the most effective topical strategy.

Bottom line: for menopausal skin, look for moisturisers containing all three physiological lipids, ceramides, cholesterol, and fatty acids, with cholesterol explicitly on the INCI list, ideally in the first 10 ingredients. Cholesterol should be the dominant lipid by ratio. The Zettersten study used a 1:3:1 molar ratio (ceramide: cholesterol:fatty acid).

The closest retail approximation is SkinCeuticals Triple Lipid Restore at 2% ceramides : 4% cholesterol : 2% fatty acids by weight, a 1:2:1 ratio, so cholesterol-dominant but at a lower fold-dominance than the study. The principle holds: cholesterol should be the largest of the three lipid components.

I have compiled a list of ‘cholesterol-dominant moisturisers’ I could find, and I will add more as I come across anything worth including.

Every other active you apply has to come back to a barrier that can hold itself together.

This is the foundation.

---

Step 2: Niacinamide (4–5%)

If menopausal skin had a single best all-purpose active, it would be niacinamide.

The evidence breadth is genuinely unusual. Most cosmetic actives have one mechanism and one outcome. Niacinamide has three, each independently studied, each directly relevant to what’s happening in menopausal skin.

Start with lipid synthesis. Tanno et al. (2000) showed that nicotinamide upregulated ceramide biosynthesis in cultured keratinocytes by 4.1–5.5 fold, with free fatty acid synthesis up 2.3-fold and cholesterol synthesis up 1.5-fold. Topical application reduced TEWL in dry skin. Those are the same three lipid classes you’re trying to rebuild from Step 1, except this time, your own keratinocytes are making them. Niacinamide and a cholesterol-dominant moisturiser are doing the same job from two different angles.

Then there’s pigmentation. Hakozaki et al. (2002) demonstrated that niacinamide inhibits melanosome transfer from melanocytes to keratinocytes by 35–68%. This matters in menopause because oestrogen withdrawal alters melanocyte regulation, melasma frequently worsens and solar lentigines accumulate. Most pigment-targeting actives (kojic acid, alpha-arbutin, hydroquinone) work upstream by inhibiting tyrosinase and reducing melanin production. Niacinamide works downstream by blocking melanin transfer. Different leverage point, same problem. They stack well.

And then the clinical endpoints. Bissett, Oblong, and Berge (2005) ran a 12-week, vehicle-controlled, split-face study in 50 women using 5% niacinamide. The results showed significant reductions in fine lines and wrinkles, hyperpigmentation, red blotchiness, yellowing, and improved elasticity. Modest effect sizes, but consistent across multiple endpoints, which is exactly what you want from a multi-mechanism active.

What to look for: 4–5% niacinamide, if you like a higher concentration, it is fine. Although higher doesn’t reliably do more. If you sting easily, scale back, menopausal barriers are more reactive than they were in your twenties.

Here is the list of my recommended serums with niacinamide.

---

Step 3: A retinoid

Retinoids are THE ONLY topical class with peer-reviewed evidence for restoring collagen production in chronologically aged skin, not just photoaged skin.

That distinction matters in menopause.

The collagen loss you’re seeing isn’t only about sun damage. It’s about hormonal withdrawal acting on dermal fibroblasts directly.

The evidence:

• Kligman, Dogadkina, and Lavker (1993), tretinoin applied to sun-protected (intrinsically aged) skin: clear histological improvement.

• Griffiths et al. (1993), tretinoin restores collagen formation in photodamaged skin.

• Kafi et al. (2007), 0.4% retinol on intrinsically aged forearm skin in elderly subjects: significant clinical and histological improvement.

The mechanism is direct. Retinoids bind nuclear retinoic acid receptors (RAR/RXR), upregulate procollagen I and III gene expression, downregulate matrix metalloproteinases (the enzymes that degrade collagen), and accelerate epidermal turnover. They are doing real work at the level of the deficits oestrogen used to compensate for.

A quick word on which retinoid to choose, because the category is bigger than people realise.

Tretinoin (prescription, 0.025–0.05%) is the gold standard, strongest evidence, four decades of clinical data, effective on both intrinsically aged and photoaged skin.

Retinaldehyde (OTC, 0.05–0.24%) sits one enzymatic step away from tretinoin: faster-acting than retinol, gentler than tretinoin.

Retinol (OTC, 0.1–1%) is two steps from tretinoin and has its own strong evidence base. The Kafi 2007 study used 0.4% retinol on intrinsically aged forearm skin in elderly subjects with significant clinical and histological improvement.

Adapalene is interesting in this context: the 0.3% prescription strength was non-inferior to 0.05% tretinoin for photoageing in a 24-week head-to-head (Bagatin et al., 2018), often better tolerated than tretinoin, and particularly useful if you also have adult-onset perimenopausal acne.

The 0.1% OTC adapalene in the UK (Differin) has strong evidence for acne but less direct anti-ageing trial data.

A word on hydroxypinacolone retinoate (HPR), sometimes sold as "granactive retinoid", because it's popular in OTC formulations right now, and the marketing claims around it deserve scrutiny. HPR is a retinoid ester of all-trans retinoic acid, and the theoretical appeal is that, unlike retinol or retinaldehyde, it doesn't require enzymatic conversion to bind retinoic acid receptors directly. That sounds biologically elegant, and brands run with it ("works like tretinoin without the irritation").

The clinical reality is more modest. The handful of published studies on HPR are small, often industry-funded, frequently use HPR in combination with other retinoids (which makes isolating its contribution difficult), and show effect sizes well below those of tretinoin or retinaldehyde.

There is no head-to-head RCT comparing HPR to tretinoin on equivalent endpoints. HPR is also stable, well-tolerated, and reasonable as a gentle introductory retinoid for very sensitive or perimenopausal-reactive skin, but positioning it as a "tretinoin alternative" overstates what the evidence currently supports.

At the bottom of the hierarchy sit retinyl esters (retinyl palmitate, retinyl acetate), they require multiple conversions in the skin and have the weakest evidence. Tolerable but underwhelming.

Whichever retinoid you choose, start lower than you think. 0.025% retinol or 0.05% retinaldehyde, two to three nights a week, building over 12 weeks. Use the sandwich method (moisturiser before and after) for the first month. And pair it with your cholesterol-dominant moisturiser. The retinoid does the dermal work, and the moisturiser holds the epidermis together while it does.

I have an entire TikTok playlist dedicated to how to start using retinoids to reduce side effects.

If retinoids are still too much for your skin, which is common in perimenopausal skin, especially in the first 12 weeks, there are two evidence-based options before you give up entirely.

The first is pairing the retinoid with topical ectoine. Kauth et al. (2022) systematically reviewed ectoine for inflammatory skin conditions, including retinoid dermatitis, and at 5.5–7% concentrations, it was as effective as dexpanthenol at reducing retinoid-induced irritation.

Mechanistically, ectoine is an osmolyte: it stabilises the lipid matrix, reduces TEWL, and exerts broad anti-inflammatory effects on disrupted skin. Cosmetic formulations at 2% and above are widely used and reasonable, though the strongest published clinical evidence sits at 5.5–7%. Apply ectoine and a ceramide/cholesterol cream as the sandwich layers around your retinoid, barrier support from both lipid replenishment and osmolyte stabilisation simultaneously.

The second option is switching to bakuchiol, but here the framing matters. Bakuchiol is often positioned as a retinol equivalent. The clinical evidence DOES NOT support that framing. The most-cited trial (Dhaliwal et al., 2019) was designed as a superiority study with a null result and has been formally critiqued for methodology in the Journal of Cosmetic Dermatology.

The 2024 systematic review of anti-ageing cosmeceuticals graded retinol an “A” and bakuchiol a “C”. Bakuchiol has decent in vitro data, antioxidant activity, fibroblast stimulation, fibronectin and collagen induction (Bluemke et al., 2022), and is well-tolerated, so it’s a defensible option if you genuinely cannot use any retinoid (pregnancy, breastfeeding, severe intolerance).

For someone seeking the dermal remodelling effects of a true retinoid, it is not equivalent.

---

Step 4: Vitamin C

Vitamin C earns its place for two reasons.

The first is structural: it’s an obligate co-factor for the enzymes (prolyl and lysyl hydroxylases) that crosslink collagen. Without sufficient ascorbate, the collagen your fibroblasts make is structurally weaker, regardless of how much retinoid stimulus you’re providing.

The second is defensive: oestrogen withdrawal reduces endogenous antioxidant capacity (Bottai et al., 2013), and topical vitamin C partially offsets that.

Lin et al. (2005) showed that combining 15% L-ascorbic acid with 1% α-tocopherol and 0.5% ferulic acid roughly doubled the photoprotective effect of vitamin C alone. That C/E/F combination is the most evidence-based vitamin C formulation, and for menopausal skin it does double duty, collagen co-factor and defence against the oxidative stress that accelerates in oestrogen’s absence.

What to look for: 10–20% L-ascorbic acid at pH below 3.5, the window where it penetrates and stays stable. Ideally, with vitamin E and ferulic acid. Opaque packaging.

Here is a shopping list for the most reliable vitamin C serums.

Vitamin C oxidises rapidly. Yellow-orange serum is partially oxidised. Deep brown is dead.

If L-ascorbic acid is too irritating, ethyl ascorbic acid or TDH ascorbate are reasonable alternatives, comparative efficacy data is weaker.

---

Step 5: SPF 50+

Here’s the bottom line on sunscreen. It is the single most evidenced anti-ageing topical you can use.

Hughes et al. (2013) ran the only long-term RCT comparing daily versus discretionary sunscreen use over 4.5 years. Daily SPF 15 reduced visible photo-ageing by 24%.

Randomised. Controlled. Intention-to-treat.

There is no other anti-ageing topical with that calibre of evidence.

In menopausal skin specifically, SPF matters more than ever, for two related reasons. Melasma and lentigines worsen with hormonal change, and visible light plus UV are the primary drivers of both.

But beyond pigmentation, there’s a deeper biological point: Rittié et al. (2008) showed that even oestradiol, applied topically in a controlled study, could not restore collagen in photodamaged skin. Only in sun-protected skin. Cumulative UV damage closes the door on a lot of restorative biology.

Sunscreen keeps that door open.

A quick mechanistic aside, because this matters for menopausal skin specifically:

UVA isn’t a monolith, it’s a spectrum. And the deep end of that spectrum, long-UVA1 band (roughly 370–400nm within the broader IVA1 range of 340-400 nm), is the wavelength range that penetrates furthest into the dermis. It passes through the epidermis and deposits a substantial portion of energy in the structural layer where collagen and elastin actually live. Once there, it activates matrix metalloproteinases (the enzymes that chew through collagen), upregulates melanogenesis, and drives the dermal damage we register clinically as photoaging.

Now overlay that on the menopausal timeline. In the first five years post-menopause, the dermis loses roughly 30% of its collagen as oestrogen support to fibroblast function falls off a cliff.

Translation: the collagen pool is shrinking from the inside, while long-UVA1 is degrading what’s left from the outside. Two assaults on the same tissue. The effect is compounding, not merely additive.

Here’s the catch most SPF marketing won’t mention: traditional UVA filters typically taper off around 370nm, exactly where the deepest-penetrating UVA1 band really starts to matter. Genuine coverage of that long-UVA1 tail only became possible with the newer generation of filters. The cleanest example is Mexoryl 400 in L’Oréal-group sunscreens, which peaks at 385nm (precisely where you want it).

I have compiled a shopping list of THE MOST comprehensive sunscreens for menopausal skin that I could find. There is just one outstanding Swiss sunscreen I was unable to include, as it is usually sold through pharmacies: Louis Widmer All Day SPF 50+ (perfumed and unscented). If you can find it in Europe, I highly recommend trying it.

The Korean alternative is filter-stacking: Uvinul A Plus + Tinosorb S + Mexoryl SX in the better formulas, which gets you meaningfully into that range, even if it doesn’t replicate Mexoryl 400’s absorption peak or fully flatten the 380-400 nm gap.

House of Hur Weightless Sun Fluid SPF50+ PA++++ and VT Mushroom SPF 50+ PA++++ Airy Sun Cream are currently the most filter-rich examples of this Korean approach.

Bottom line: in menopause, SPF stops being just about sunburn prevention. It’s about defending a dermis that’s already losing structure. Choose filters that cover the wavelengths actually doing the damage, including the long-UVA1 tail.

And remember, the “best sunscreen” is the one you’ll wear daily.

---

Optional supporting players

These have evidence, but rank them below the five core steps.

• Peptides: pal-KTTKS (Matrixyl) and GHK-Cu have small but reasonable RCTs for collagen-related improvement (Robinson et al., 2005; Pickart and Margolina, 2018). Useful as adjuncts; not a retinoid replacement.

• Phytoestrogens (genistein, isoflavones), Silva et al. (2017) compared 0.01% topical oestradiol with 4% genistein over 24 weeks. Both increased Type I and III collagen. Oestradiol was superior. Genistein had measurable but modest effects. They are not “natural HRT.” They are weak ER-binding plant compounds with small effect sizes.

• AHAs / PHAs, lactic acid 5–10% or gluconolactone 8–10%, once or twice a week max. Helpful for surface texture. Skip if your barrier is reactive.

---

Your actual daily routine

Morning: Gentle cleanser ( optional )→ Vitamin C serum (or C/E/F) → Niacinamide 4–5% → Optional: peptide serum → Cholesterol-dominant moisturiser → SPF 50+.

Evening: Gentle cleanser → Niacinamide 4–5% → Optional: peptide serum →Retinoid (start 2–3 nights/week, build up) → Cholesterol-dominant moisturiser (sandwich method around the retinoid as needed).

Weekly: Optional AHA/PHA, 1–2 nights, swapped in for the retinoid or in the morning, first thing, if your skin can tolerate.

That’s it.

Not 14 steps.

---

On topical oestrogen: a biochemist’s honest answer

This question is everywhere right now. A lot of online recommendations are dangerously casual about it.

So let me lay this out plainly.

Does topical oestrogen work on menopausal skin?

Answer: yes, partially, in some studies, with caveats.

Schmidt et al. (1994) and Varila et al. (1995) showed improvements in collagen, wrinkle depth, and elasticity at 3–6 months. Punnonen et al. (1987) showed elastic fibre improvement with topical oestriol.

But Rittié et al. (2008) showed oestradiol stimulated collagen on sun-protected skin but had no effect on photodamaged forearm or facial skin. That’s a significant limitation that gets glossed over.

Now the part that really gets glossed over.

Oestradiol has a molecular weight of 272 Da. Oestriol is 288 Da.

Oestradiol and oestriol are uncharged, lipophilic steroids, under 500 Da in molecular weight, ideal transdermal candidates.

That’s not coincidence. That’s exactly why oestradiol patches and gels (Estraderm, Sandrena, Oestrogel) work as licensed systemic HRT. They were designed to absorb into the bloodstream through the skin.

A “topical” oestrogen on your face is not pharmacologically the same category as a topical vitamin C serum.

It is a steroid hormone that absorbs systemically, by molecular weight, by lipophilicity, by design.

The numbers back this up.

In the same Rittié 2008 study, the pharmacokinetic arm tested topical oestradiol across a range of concentrations (0.01% to 2.5%) and measured a substantial dose-dependent rise in serum oestradiol in postmenopausal women at the higher concentrations tested.

That is not a trace level at the higher doses. That is meaningful systemic exposure from a “facial cream.”

Older small studies (Kainz et al., 1993) reported no detectable systemic change with very low doses (0.01% oestradiol or 0.3% oestriol). Their detection limits were modest and their cohorts small.

The reassurance that “topical oestrogen doesn’t absorb” doesn’t hold up at the concentrations being recommended online.

Oestriol vs oestradiol.

Oestriol binds oestrogen receptors more weakly than oestradiol and shows preferential affinity for ERβ (Kuiper et al., 1997; Zhu et al., 2006). It is less systemically potent than oestradiol. But it is not non-oestrogenic. Calling oestriol “safer” overstates the case. Calling it “weaker” is fair.

Regulatory status in the UK and EU.

All oestrogens, oestradiol, oestriol, and conjugated equine oestrogens are listed in Annex II of EU/UK Regulation 1223/2009 as substances prohibited as cosmetic ingredients.

They are prescription-only medicines (POM) under the Human Medicines Regulations 2012. A facial oestrogen cream sold or compounded as “skincare” is, by definition, either an unlicensed medicine or a non-compliant cosmetic.

Neither is a regulatory grey area.

Both are unlawful in the UK.

Contraindications still apply.

History of breast, endometrial, or ovarian cancer. Undiagnosed vaginal bleeding. Active liver disease. Previous VTE or stroke.

When the molecule is on your face, it’s still the molecule.

If you’re considering compounded topical oestrogen for skin reasons, that’s a conversation with a menopause specialist or dermatologist who knows your full history, not a TikTok recommendation, not a cosmetic-counter purchase.

To be clear: I am not anti-HRT.

HRT is one of the most evidence-based interventions in menopause medicine. Many women see meaningful skin improvement from systemic HRT taken for the right indications.

That is a clinical decision with a prescriber. Not a skincare swap.

The line I draw as a cosmetic chemist is this: facial oestrogen creams marketed as skincare are mislabelled medicines, and the safety profile is being underplayed.

---

What won’t fix menopausal skin

None of them work.

“Collagen creams.”

Topical collagen sits on the surface as a humectant. It does not become your collagen. If a product claims to “replace lost collagen”, it doesn’t. Look for ingredients that stimulate your own fibroblasts (retinoids, vitamin C, peptides) instead.

“Stem cell” creams/serums.

There are no living stem cells in your moisturiser or serum. What you have is cell extract or conditioned media, protein-and-lipid mixtures with poorly characterised activity. The marketing claim is much bigger than the data.

Exosome serums.

The buzzword of the last two years. Exosomes are real biology, small (30–150 nm) extracellular vesicles that carry mRNA, microRNA, and signalling proteins between cells. In medicine, autologous human-derived exosomes are being studied seriously for wound healing and regeneration.

The cosmetic category is not that.

Almost all “exosome” topicals sold over-the-counter are plant-derived (rose stem cell, lotus, ginseng, snow mushroom) or bacterial lysate-derived. Plant exosomes do not signal to human cells the way human exosomes do. The receptor recognition, the cargo content, and the membrane composition are species-specific.

The few topicals claiming “human-derived exosomes” raise a separate problem: they’re either from cultured human cell lines (poorly characterised cargo, limited stability data), or they’re regulatorily ambiguous, in the UK and EU, products containing human-derived biological material face medicinal-product classification, not cosmetic.

Stability is also a real issue. Exosomes degrade rapidly outside controlled conditions; a jar sitting at room temperature in your bathroom is not a controlled condition.

Bottom line: this is one of the most overpromised categories in skincare right now. The biology is real. The retail products mostly aren’t.

Generic “ceramide creams” without cholesterol.

Useful, but, as covered in Step 1, they don’t address the rate-limiting lipid deficit in ageing skin. Look for the full triplet.

Phytoestrogens marketed as “natural HRT.”

They aren’t. They’re weak ER-binding plant molecules with modest topical effects. Use them as adjuncts, not foundations.

DIY oestrogen creams.

See the section above.

Don’t.

---

How long does this actually take?

I’ll be straight with you. Skin remodels slowly. Menopausal skin remodels more slowly.

• Barrier improvements (dryness, tightness, reactivity): typically 2–6 weeks once a cholesterol-dominant moisturiser and niacinamide are in place.

• Pigment changes (lentigines softening, melasma calmer): 8–16 weeks with consistent SPF + niacinamide + tyrosinase inhibitor ± retinoid.

• Visible collagen-related changes (line softening, firmness): 12–24 weeks minimum with retinoid use. Longer for meaningful results. Six months is the minimum for honest before-and-after. Twelve months is more realistic for the bigger structural changes.

This is a long game.

It’s not a hopeless one.

---

The bottom line

Menopausal skin is a biology change. Not a failure. Oestrogen withdrawal removes a regulatory signal that supported collagen synthesis, lipid barrier, antioxidant defence, and melanocyte regulation, all at once. The right protocol addresses the downstream consequences in the right compartments.

Your five steps:

1. A cholesterol-dominant moisturiser: the most under-discussed, best-evidenced swap for ageing skin.

2. Niacinamide 4–5%: ceramide synthesis, melanosome transfer, fine lines and redness in one molecule.

3. A retinoid (tretinoin, retinaldehyde, retinol, or adapalene), for the dermal collagen work.

Bakuchiol only if you genuinely cannot use any retinoid.

4. Vitamin C, collagen co-factor and antioxidant defence, ideally as C/E/F in opaque packaging.

5. SPF 50+ daily, the only topical with long-term RCT-grade anti-ageing evidence.

Be sceptical of: “collagen creams,” generic ceramide-only moisturisers, “stem cell” cosmetics, exosome promises, phytoestrogens marketed as natural HRT, and anything claiming to “replace lost oestrogen” topically.

Do not use compounded facial oestrogen creams as cosmetics. They are systemically absorbing prescription medicines, prohibited as cosmetic ingredients in the UK and EU. The safety profile demands a prescriber’s involvement, not a serum recommendation.

Give this protocol six months.

Expect barrier and pigment changes earlier. Expect collagen-driven changes to take longer.

Be kind to your skin while it adjusts.

And if you’re considering systemic HRT for menopausal symptoms more broadly. That’s a conversation with a menopause specialist. One that often improves skin alongside everything else.

You’ve got this. x

---

Marina is a cosmetic chemist, biochemist, and science-based skincare creator and educator. Follow her on TikTok, Instagram and YouTube @4complexion for evidence-based skincare content.

Disclaimer: This article is for educational purposes only and is not medical advice. I’m a cosmetic chemist, not a doctor or a menopause specialist. Decisions about HRT, prescription topicals, or any concerns about persistent skin or hormonal changes should be made with your GP, dermatologist, or menopause specialist.