In partnership with
You Don't Need to Read the Whole Book
How can you change your life? In Atomic Habits, James Clear argues that the key lies in your habits: the automatic behaviors that make up more than half of what you do every day.
Shortform gives you that kind of summary for thousands of nonfiction books: the key ideas, the reasoning behind them, and how to put them to work, in about 15 minutes.
Then do the same for the next book on your list.
LONGEVITY LATEST ISSUE 34 · 28 OCTOBER 2026
LONGEVITY LATEST
The Evidence-Based Edge on Living Longer and Better
Issue 34 · The Collagen Question · 28 October 2026
Welcome
Twenty-four per cent. That’s how much less skin ageing showed up over four and a half years in the Australians told to wear sunscreen every morning, compared with those left to use it when they remembered. The daily group showed no detectable increase in skin ageing at all. It remains one of the cleanest anti-ageing results from any randomised trial, and the product is in every supermarket.
The market has since moved on to peptides: collagen powder in the morning coffee, peptide serums, and injectable “research” vials passed around gyms and podcasts. This week I graded them against the older tools with real trials behind them, and held a set of weights to the same standard. (Eyes, promised for this week, move to later in the autumn.)
SPOTLIGHT
Six popular skin fixes, graded
What you’d use | What the best evidence shows | Grade |
Daily sunscreen | Nambour trial: 903 adults under 55, four and a half years. 24% less skin ageing than occasional use. | A · prevention |
Tretinoin cream (prescription) | 8 randomised trials, 1,361 people: fine and coarse wrinkles improve against a plain cream. | A · repair |
Collagen powder | Small-to-moderate gains in hydration and elasticity; nothing reliable for wrinkles. Reviews disagree on whether funding explains the gains. | C · optional |
Strength training | One trial, 56 women: elasticity improved, but equally with cycling and with no inactive control. Thickness change was tiny. | C · watch |
Peptide serums (Matrixyl, copper peptides) | Small, short trials, many run or written up by the companies and inventors behind them. | C · unproven |
Injectable BPC-157 | No published randomised human trial. Two tiny uncontrolled reports; nothing for skin or ageing. | F · avoid |
My reading: the two A grades are the oldest, cheapest options on the list. The newest and most talked-about has the least behind it.
Hughes et al. Ann Intern Med (2013) · Huang & Lee. Dermatol Pract Concept (2025) · Myung & Park. Am J Med (2025) · Batool et al. J Cosmet Dermatol (2026) · Nishikori et al. Sci Rep (2023) · Vasireddi et al. HSS J (2025)
THIS WEEK’S ANALYSIS
Top 3 Interventions Under the Microscope
Skin-evidence grades judge visible skin ageing only: A = consistent randomised trials show visible benefit; B = positive randomised evidence, but few, small or short trials; C = mixed or preliminary evidence, or benefit mainly in industry-linked trials; D = mostly lab and animal data; F = claims far outrun the data.
1. Tretinoin, for wrinkles | Grade A
The benchmark. If a product claims to beat it, ask to see the head-to-head trial.
How it works. Tretinoin is vitamin A in its active acid form. It acts directly on genes in skin cells and pushes fibroblasts in the dermis to make new collagen while holding back the enzymes that cut it.
The evidence. A 2025 meta-analysis pooled 8 randomised trials with 1,361 people: fine and coarse wrinkles both improved significantly against a plain cream. The biopsies show why. In a 1993 NEJM trial of 29 people with sun-damaged skin, new collagen formation rose 80% after 10 to 12 months of 0.1% tretinoin and fell 14% on the plain cream. Collagen formation isn’t the same as a smoother face, but here the two point the same way.
Fair warning. The first month is usually the worst: redness, flaking, stinging. Start two or three nights a week, pea-sized, with moisturiser on top, and pair it with daily sunscreen. Not in pregnancy or when trying to conceive. In the UK it’s prescription-only; over-the-counter retinol is gentler, with less evidence. US online services charge about $30 a month.
Huang & Lee. Dermatol Pract Concept (2025) · Griffiths et al. NEJM (1993) · Nurx. Tretinoin pricing (2026)
2. Collagen powder, for wrinkles | Grade C
Some help for hydration and bounce; nothing reliable for wrinkles.
The idea. Eating collagen to build collagen is a bit like posting someone a box of loose bricks and expecting your house to come back. Digestion breaks most of it down. The better argument is that a few fragments reach the blood intact and may signal to the cells that make collagen.
The trials. Two recent meta-analyses agree on wrinkles and disagree on almost everything else. The 2025 review in the American Journal of Medicine (23 trials, 1,474 people) was positive overall, but found no benefit in trials without industry funding and no significant effect in the high-quality ones. This year’s update in the Journal of Cosmetic Dermatology found small-to-moderate gains in hydration and elasticity, and reports that independently run trials showed effects similar to sponsored ones. On wrinkles, both land in the same place: the 2026 estimate is −0.02, with a range that includes no effect at all.
Why C? A real but modest effect on how skin feels is plausible. The wrinkle claim on the tub isn’t supported. A 567 g tub of a leading brand costs £33–50 in UK shops and lasts about four weeks at the label serving. Fine as a protein top-up; don’t pay extra for the wrinkle promise.
Myung & Park. Am J Med (2025) · Batool et al. J Cosmet Dermatol (2026) · Medino. UK price comparison (2025)
3. Strength training, for skin | Grade C
Interesting, early and over-reported. Train for the many reasons that are proven; treat any skin benefit as unconfirmed.
The trial. Japanese researchers randomised 61 sedentary women aged 41 to 59 to 16 weeks of supervised cycling or machine weights, twice a week; 56 finished. Elasticity rose in both groups (0.32 to 0.37 on the device’s recovery scale) and the upper dermis looked denser on ultrasound. Those changes were clear.
The headline claim, that weights thickened the dermis, is much softer than it sounds. In the weights group thickness went from 1.77 to 1.79 mm, not significant on its own (p = 0.20); the paper’s case rests on a between-group comparison it calls marginal. I wanted to like this one. I can’t give it a B.
What holds it back. No group stayed sedentary, so the trial can’t separate exercise from season, measurement drift or simply being in a study. Three of the authors work for POLA, a Japanese cosmetics company. That doesn’t make the result wrong, but it gets the same disclosure I’d give a collagen trial. Lift for your muscles, bones and metabolism. If your skin benefits too, that’s a bonus nobody has proven yet.
HYPE CHECK
“The Wolverine peptide”
The claim. BPC-157, a 15-amino-acid fragment of a protein first found in gastric juice, is sold online for healing and rejuvenation: tendons, gut, joints, skin. “Research” vendors list 10 mg vials at roughly $25–50, labelled not for human consumption. This summer’s US regulatory news has been read as a green light.
The evidence is thin rather than absent. A 2025 systematic review, covering studies to June 2024, found 36: 35 in animals or cells, and one retrospective report of 12 people given a knee injection for chronic pain, with no comparison group. Since then, a two-person intravenous safety pilot reported no adverse effects. A phase I trial registered in 2015 never reported results. There is still no published randomised trial in humans, and nothing at all for skin or ageing.
Two uncontrolled human reports, 14 people in total. No published randomised trial.
Why it misleads. The headlines are about permission, not proof. FDA’s April change came after nominators withdrew, not after a safety finding, and July’s narrow advisory vote to allow compounding went against FDA’s own reviewers and isn’t binding. Product quality is a gamble too: of the 200 BPC-157 tests shown on one independent lab’s public tracker, 91 failed its purity or dose standard. That’s one tester’s sample, not the whole market, but it’s not reassuring.
Vasireddi et al. HSS J (2025) · Lee & Burgess. Altern Ther Health Med (2025) · D.J. Holt Law. FDA peptide update (2026) · Finnrick. BPC-157 test tracker (2026) · Peptide Catalog. Vendor prices (2026)
Our verdict. Avoid. Injecting an unregulated vial on the strength of rat studies is an uncontrolled experiment with you as the only subject. One vial costs about what a month of US prescription tretinoin does, and tretinoin has decades of randomised trials behind it.
IN BRIEF
Semaglutide and the ageing clock. A placebo-controlled trial in 108 adults with HIV-related abdominal fat, published in Nature Communications in May, found a 9% slower pace of ageing on the DunedinPACE epigenetic clock after 32 weeks, among the 84 with paired blood samples. The analysis was post hoc, older clocks barely moved, and a clock is a measurement, not a health outcome. The authors say plainly it isn’t an anti-ageing drug.
THIS WEEK’S DEEP DIVE
The Scaffolding Problem
Take collagen-making cells out of the skin of people over 80, grow them in a dish, and they still make about two-thirds as much collagen as cells from people in their twenties.
So why does ageing skin make so little? The companion explains how those cells lose their grip, why that makes retinoids more plausible than most alternatives, and why molecular size makes many peptide serums a long shot.
BIOHACKING CORNER · THE PLAN LIVES HERE
Morning: sunscreen, every day. Broad-spectrum SPF 30 or higher with a 4- or 5-star UVA rating, as the NHS advises, on face, neck and the backs of your hands. Reapply every two hours outdoors. The Nambour volunteers used SPF 15+ in the 1990s; the daily habit did the work.
Night: a retinoid, slowly. Pea-sized, two or three nights a week for the first month, then nightly if your skin allows. Dry skin first, moisturiser on top.
Twice a week: lift. Six machine exercises covering legs, arms, back, shoulders and chest, three sets of ten, building to a load you can only just manage for ten reps. Do it for strength and metabolic health; count any skin change as unproven.
My rule, a judgement rather than a finding: give any skin routine 12 weeks, with photos under the same light on day one and day 84, before deciding whether it works.
CLOSING
Protect the scaffold, then rebuild it
The tools that slow skin ageing aren’t secret; they’re just less exciting than an injectable with a superhero nickname. Next week: VO2 max, the fitness measure with some of the strongest links to how long people live, and the interval sessions that raise it fastest.
Stay curious and stay healthy!
Christian Thomsen, Editor
Longevity Latest is published weekly by FrontWave Media Ltd. Educational content, not personal medical advice. Retinoids, including tretinoin and retinol, should not be used in pregnancy or when trying to conceive; ask a pharmacist or GP before starting one. BPC-157 is not approved for human use in the UK, EU or US. Check with your doctor before starting a new training programme if you have a heart, joint or other health condition, and before taking supplements alongside prescribed medicines.
Reading notes
Evidence checked through 24 October 2026; each claim links to its source beside it. Nambour used SPF 15+ between 1992 and 1996, with skin ageing graded blind. The strength-training trial compared two exercise types with no inactive control; dermal thickness was not significant within groups. The two collagen reviews differ on funding and quality; both find no reliable wrinkle effect. BPC-157 human data are one retrospective series (to June 2024) and one two-person pilot (2025); FDA status reflects the April and July 2026 actions, with no final rule. Finnrick figures cover the 200 tests on its public history, accessed 24 October 2026. The semaglutide clock analysis was post hoc. Prices: UK retail (October 2025) and US online listings (October 2026).
© 2026 FrontWave Media Ltd · Longevity Latest
Stop typing what you could say in 10 seconds.
Wispr Flow turns your voice into clean, professional text inside any app. Emails, Slack, client updates: speak once, send without editing. 4x faster than typing. Works on Mac, Windows, iPhone, and Android.